the study of managerial behavior || studies of managerial jobs: methodologies and profiles

STUDIES OF MANAGERIAL JOBS: METHODOLOGIES AND PROFILESAuthor(s): Rosemary StewartSource: International Studies of Management & Organization, Vol. 2, No. 1, THE STUDY OFMANAGERIAL BEHAVIOR (SPRING 1972), pp. 7-37Published by: M.E. Sharpe, Inc.Stable URL: http://www.jstor.org/stable/41103783 .

Accessed: 14/06/2014 03:58

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

M.E. Sharpe, Inc. is collaborating with JSTOR to digitize, preserve and extend access to International Studiesof Management &Organization.

http://www.jstor.org

This content downloaded from 195.78.108.199 on Sat, 14 Jun 2014 03:58:09 AMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=mes

http://www.jstor.org/stable/41103783?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


STUDIES OF MANAGERIAL JOBS: METHODOLOGIES AND PROFILES*

Rosemary Stewart (UK)

I. LOOKING AT MANAGERST JOBS

1. Four Ways of Looking

What do we know about managers' jobs? The answer is "very little" - as the author found when she tried to write a chapter on the manager's job in her last book. (1) The idea for the research described in this book developed as a result of the problems of trying to write that chapter. A study of the literature showed that we know very little about what managers do.

There have been four main approaches to the problem of analyzing managers' jobs. These are represented by four questions: "What is management?"; "What are the responsibilities belonging to the different jobs that make up an organization?";

*This is an extract from Chapter 1 ("Looking at Managers' Jobs") and a condensed version of Chapter 6 ("Job Profiles") of Rosemary Stewart's book Managers and Their Jobs; A Study of the Similarities and Differences in the Ways Managers Spend Their Time, London and Basingstoke, Macmillan, 1967. The material is reprinted with the permission of the publisher. Dr. Stewart has appended a short commentary on the present situation, including some observations on the other articles in the present issue.

7



8 Rosemary Stewart (UK)

"How can management jobs be compared and evaluated?"; and "How do managers spend their time?" The questions differ because the questioners have different aims.

The first question, "What is management?", is discussed in many management textbooks. It can be called the traditional or classical approach. The classical writers, such as Fayol and Urwick, looked at management as a unitary activity and asked themselves, "What are the tasks that all managers perform?" The classical writers had some disagreements about terminol- ogy, but in general they agreed that managers planned, orga- nized, motivated, and controlled. This kind of analysis was useful in giving some idea of the nature of the manager's job, but was too general to be of much help in deciding how managers should be selected and trained. Nor did it make much contribu- tion to the problems of planning an organization.

The second approach to an analysis of managers' jobs seeks to describe the responsibilities of all the supervisory and managerial jobs in an organization. It does this by preparing individual job descriptions. These are based on questions about the job that are answered by its occupant and usually by his superior as well. Considerable work has been done to refine job descriptions by developing more detailed and penetrating questions to ask about the job. Job descriptions are a useful tool for management. They can help in preparing a job specification of the qualifications and experience required for a particular job. They may also be an aid to organization planning. But they have limitations. One is that they may be descriptions either of what is supposed to happen, or of what those concerned think happens, not necessarily of what does happen. Another limitation is that they apply best to a fairly static organization and least well to one that has to adapt itself to changing conditions, where job responsibilities may themselves be changing. Chappie and Say le s have tried to overcome the first limitation by pro- ducing a different sort of job description, which aims to describe the job in terms of work flow. (2) They say that a job description should tell us what the manager does, with whom, when, where, and how often, instead of using vague phrases such



Studies of Managerial Jobs 9

as "work with.n This is a combination of the second and the fourth approach to management, which seeks to apply the information collected by asking MWhat do managers do?" to a description of the individual job.

The third approach to looking at managers1 jobs asks "How can they be compared and evaluated?" The usual form of job description does not provide an adequate basis for comparing one job with another for salary purposes. It is too descriptive and not sufficiently analytical to provide a good basis for comparisons between jobs in very different departments. Job evaluation as applied to manual and junior clerical jobs cannot be used for the more complicated tasks performed by managers. Two different approaches to the problem of finding a basis for objective comparisons between different types of managerial jobs have been used by Doulton and Hay (3) and by Jaques. (4) Doulton and Hay developed their method initially in order to devise salary grades for the widely differing types of managerial and specialist jobs found in the British Broadcasting Cor- poration. Their work had affinities with job descriptions and job evaluation. They sought to assess all aspects of a job under the following five main headings: application of knowledge and experience; judgment; creative thought; man management (which was defined to include dealing with people who are not one* s subordinates); and decisions.

Doulton and Hay compared managerial or professional jobs under a number of different headings. Elliott Jaques approached the same problem of comparison by asking whether there was one aspect of managers' jobs that could be used to assess the relative value of one job compared with another. After an ex- tensive comparison of jobs in one manufacturing company, Glacier Metal, he decided that the answer was the time- span of discretion in a job. Differences between jobs could be measured by comparing "the maximum period of time during which the use of discretion is authorized and expected, without review of that discretion by a superior." (5) Jaques' s approach, unlike the others that have been described, sought to measure one manager's job against another. In this respect he resembles




those who have tried to answer the next question. The fourth question that has been asked about manager sT jobs

is "How do managers spend their time?" It is the question asked in this study. People who ask this question think that we need to know more about management in practice in order to try to improve our theories of management. Little of the thinking about management is based on objective data. There have been few attempts to collect and classify data about management behavior, either to test hypotheses or to provide information for the formulation of hypotheses.

2. How Do Managers Spend Their Time ?

What questions should one ask? The question "What does a manager do?" is not an easy one

to answer. One can ask what a milkman does and then go and watch him at work. Provided that one has a logical plan for one's observation, one will come back with a clear picture of his job. We cannot do that for managers' jobs since they are both more complicated and far more varied. Hence, one has to decide what aspects of the job one wants to study. This will de- pend upon what questions one is trying to answer. The questions that one asks will also help to decide what measures one should use.

Two main measures of a manager T s activities can be used. One is the amount of time that he spends on particular activities. The other is the frequency with which he does something. For many purposes time is the more comprehensive measure, but for some, frequency may be the better measure. Sometimes a combination of the two will be desirable. One may, for instance, want to know both how long a manager spends with his subordinates and how frequently he sees them. Time is the main measure used in previous studies. It is also the main measure used in the study described in this book, but frequency is used for some aspects of the managers' work.

Let us look at some of the questions that one might ask about a particular manager's job, and at the reasons for asking them.




The list given below does not include questions that can be answered without studying what a manager does, such as "How many people does he have reporting to him?", though such questions are also important to understanding the nature of a manager's job.

a) How specialized is the job? Specialization is usually thought of in terms of subject matter: the more restricted the subject matter, the more specialized is the job. Another aspect of the amount of specialization in a job, which is rarely considered, is that of the range of contacts that it involves. (6) A job that is very specialized in its subject matter may have a wide variety of contacts.

Information on the amount of specialization could be useful for three reasons. First, it could show the range of knowledge that the job holder would need: for instance, would some knowledge of accounting be essential, useful, or irrelevant? Second, it could show whether the job was likely to be a good training ground for those senior posts that require a broad understanding of the business. Third, it would probably tell us something about the qualities of leadership needed in the job. The leadership of subordinates in a very specialized department will be based, at least in part, on superior knowledge or experience.

b) To what extent does the job involve working with other people ? All managers have to get things done through other people, but this is much more true of some jobs than of others. It would be useful for selection and training to know how important an element this is in a job.

c) What kind of contacts does the job involve? The title of the job, and its position on an organization chart, may provide no indication of the range of contacts. A job may involve contacts with people at the same level in a variety of other departments. Such a job could be said to have wide horizontal contacts. An- other job, and sometimes even the same job, may require contacts with those at several levels above and below the manager's own position in the hierarchy. Such a job can be described as having a long vertical range of contacts. Most large organizations have some jobs that provide a much greater variety of




contacts than other jobs at the same level - a fact that ambitious young men soon discover. Such jobs can be useful for training those with promotion potential. These might be called bridging jobs, as they bring their holders into contact with some of the problems and personalities in other departments and at levels of more than one remove. The management development department needs to be aware of such jobs in the organization, and of the special advantages that they can give their holders both in understanding different parts of the organization and in being noticed by senior managers outside their own departments.

A knowledge of the type and range of contacts associated with particular jobs can also be useful for training. Training in com- munications is now part of many post- experience courses, but attention is seldom paid to the different types of communication that different jobs may involve. The problems of communicating effectively will vary with the following factors: the relative levels of those concerned; the degree of common understanding of the subject; the extent to which those involved have a common background; the relationship that exists between them; and the extent to which they do or do not agree about work objec- tives.

d) What form do most of the contacts take ? Some jobs involve mainly face-to-face contacts; in others the telephone is an important means of communication. In some jobs most discussions take place between the manager and one other person; in others, the manager will spend much of his time in committees or in informal group discussions. Different forms of contact require different skills, which may have to be learned. The art of chairing a committee, for example, is one that few managers can master without experience.

e) What type of work pattern does the job tend to impose ? Jobs have different work patterns. The work pattern of a job is shown by the duration of its activities, by the extent to which its occupant can hope to plan the use of his time, by the number and frequency of interruptions, and by the extent to which activities repeat themselves. Some jobs are much more fragmented than others; some are subject to much greater time pressure




than others. One can distinguish, at one extreme, a job that per- mits the manager to plan his activities with a reasonable chance of being able to keep to his program, and, at the other extreme, a job where what he does each day is largely determined by other people. In the former he may often spend an hour or more on one activity. In the latter, the day is likely to consist mainly of short episodes. A knowledge of the pattern of a job can show some of the pressures that the job imposes upon its holder. Such information can be useful both in selection and in training. In selection it can help in eliminating those who would not stand the pressures that can be imposed by a very fragmented job, or by one subject to great time pressures. In training, such knowledge can be used to help the manager to surmount the particular pressures of his job.

f) What kind of decisions does the job involve ? This question has multiple answers, depending on what aspects of decision- making are being considered. Decisions vary in many ways: in the kind of knowledge that is required; in the amount of information that is available; in the standard of judgment that is needed; and in the degree of uncertainty that there is about the correct- ness of the decision. Decisions also vary in their importance for the success of the enterprise. The decision process may be a short-term one; many jobs only necessitate that kind of decision. The decision process may also be long-term and discon- tinuous. This is true of many of the more important decisions. Some decisions require moral courage - particularly those that affect other people.

A knowledge of the kind of decisions that have to be made in particular jobs would help in selection by showing what knowledge and experience is required, and, for some decisions, what kind of moral character. Such knowledge would also be useful for training for promotion.

g) How much time is spent on different aspects of managersT work? There is general agreement that managers are concerned with activities like planning, organizing, motivating, and controlling. Hence, jobs can be analyzed in terms of the proportion of time that is spent on each of these activities, as one guide to




their relative importance in the job. There are a few jobs that are predominantly concerned with planning. There are others where organizing is particularly important. A works manager T s job usually requires more ability to motivate other people than does that of the manager of a professional department.

Managers' jobs can also be analyzed in terms of the amount of time spent on such activities as paperwork, telephoning, discussion, and inspection. All managers' jobs involve discussions, but in some jobs they may occupy nearly all the working day. In others paperwork may take up nearly as much time. This may seem too simple an analysis to tell us anything of interest about a manager's job, but previous studies have indicated that the proportion of time that a manager spends in discussions may tell us something about the kind of situation in which he is working. One other question might also be asked.

h) How much, and what kinds of, variety are provided by the job? Jobs vary in different ways. Some have a regular monthly or seasonal cycle. Others have a similar pattern each week with only minor variations. Yet others have little routine, as the problems that they have to deal with change frequently. Differ- ent aspects of the job may vary, such as the place of work, the people contacted, and the tempo of work and its content.

A picture of the ways in which a job may vary from day to day, from week to week, or over a longer period can help to give a clearer idea of the nature of the job. It can show the kind of adaptation that is required. It can show too some of the stresses in the job as well as some of the interest. It can indi- cate whether it is the kind of job that will provide a challenge for a long time and thus help a manager to develop, or whether it is one that an able man should not hold for long because of its limited variety.

These then are the sort of questions that one needs to ask if one is seeking to study what managers do. Which of these questions have been asked, and what methods have been used to try to answer them? Let us look first at the methods available and then at the studies that have been made.




Methods of studying how managers spend their time Three main methods have been used by research workers.

The first is to ask managers to estimate how they divide their time between different activities. The second is for the manager to keep a record himself. The third is for an observer to record what the manager does. All three methods require considerable care in describing what it is that the cooperating managers are asked to do.

The first method has a number of advantages over the other two. It is much the simplest. It is easier to get managers to cooperate since cooperation will take little time. It is also a method that can be used at training courses, where there are subjects readily available. More willing cooperation makes it possible to select the sample in the expectation that most will agree to take part. The analysis of the material is much quicker as there is only one set of figures for each manager - his estimate of how he spends his time under each heading. Because of these various advantages, the method can be used much more easily to study large numbers than the other two methods, which are more costly and time-consuming.

The first method suffers from the great disadvantage that peopled estimates of how they spend their time may be wrong, and the research worker may not be able to judge how far they are wrong, or in what direction. Considerable differences between the estimated and recorded time for some activities have been shown when the two have been compared. (7) This might not matter if the errors were only individual ones, for then estimates by a number of people in similar jobs might even out individual errors. However, BurnsTs findings quoted in the note suggest that there may be general, rather than individual, ten- dencies to over- or underestimate certain aspects of the job.

The second and third methods, of self-recording in a diary and observation, are likely to be more reliable than self-estimates. They also have the advantage that they can show variations in how an individual spends his time. They have the disadvantage that they are more time-consuming for the research worker and make greater demands on the individuals being studied.




The main advantages of the diary method compared with observation are: (1) it is less time-consuming, less expensive, and much less restricted in locality - hence many more managers can be studied over a wider area of industry and locality; (2) it is easier to record the activities for a longer period, as with the observation method the longer the period of observation the fewer the number of people that can be studied; (3) classification is made by a man who knows what he is doing; for some types of analysis the observer would have to ask the manager what he was doing; (4) all time can be recorded, whereas an observer may be excluded from confidential discussions.

The main disadvantage of diaries is that they greatly limit both the scope and the content of what can be studied. The scope is limited because the manager cannot devote much time to the recording, and the content because it is difficult to get managers to record in the same way if the item being recorded allows much scope for differences in interpretation.

The main advantages of observation compared with diary keeping are: (1) the observer has time to make more detailed and comprehensive recordings; (2) the record is likely to be more complete and the observer is much less likely to omit a recording through pressure of work; (3) he can apply a consis- tent standard when recording the activities of different people.

Both self-recording and observation can either aim to cover the whole period under review or can use the incident method of recording, by sampling at random intervals. This means that the manager, or the observer, records what the manager is doing at random intervals throughout the day. The incident method is more economical of time and money than recording the whole period, whether by using a diary or by an observer. It imposes less of a burden on the manager than a continuous diary. It also spares him the constant presence of an observer. The incident method is especially useful for an observer as it enables him to study more people. It is most useful where all those to be studied are under one roof, so that the observer can keep in touch with them (although he will not be able to do that if they leave the building). The incident method limits observa-




tions to those activities that can be immediately interpreted, since the observer may find it more difficult to understand what is happening when he is only observing for brief periods. This method can only be used with self-recording if there is some device, such as a random alarm, for telling the manager when he should make a record. (8) The proportion of time spent in particular activities can be estimated by the incident method, but it cannot give information about the duration of the individual incident.

Let us look now at the studies that have been made by these different methods.

Previous studies of how managers spend their time The first reported study of how managers in industry spent

their time was made by the Swedish professor Sune Carlson. (9) He studied nine managing directors of Swedish firms for four weeks each. He mainly used the self-recording method, but checked the managing directors1 recording of contacts and phone calls by a separate record kept by their secretaries. He found out where his subjects worked, how much time they spent with other people, and who these people were. He used three classifications of the topics handled. The first was for the function, such as sales or production. (10) The second distinguished between topics that were concerned with current operations and those that were concerned with development. The third distinguished between the formulation or application of policy. The kinds of action taken were also classified under headings, such as taking decisions, giving orders, and executing.

Carlson1 s study provided the first information about what a number of industrial managers actually did. It aroused interest in this approach to studying management and became a source book for subsequent studies. Carlson gave detailed information on the methods used for recording. He did not attempt to solve the problem of getting the managing directors to classify their work in the same way so as to obtain comparable information.

Carlson was struck by the fact that the managing directors were rarely alone. Even when they were, it was usually for




periods that were too short for sustained thinking. He thought that they were too much ruled by their engagement book, so that their work was determined by other people.

Carlson suggested the idea of uadmini strati ve pathology. " (11) He thought that it was important for executive training to study administrative pathology, that is deviations from efficient procedures, and the causes of these deviations. He thought that one administrative pathology was the managing directorsT belief that their very time-consuming external activities were an abnormal call on their time. Hence they did not acknowledge the need to plan their work to allow for their frequent absences from the company. Just as many golfers have an idealized view of their normal game, so the managing directors had an idealized view of their normal work. They disregarded what they were doing at the time - and had been doing for a long period - as abnormal and temporary.

Six years later, in 1957, Tom Burns published a study of seventy- six British top managers who kept a diary for at least three, and up to five, weeks. (12) They were the top managers of six companies employing from 500 to 900 people and the top managers of one department in each of two other companies. Burns collected similar information to Carlson, but omitted two of the classifications of topics. The most interesting of his findings was the relationship he suggested between the rate of change in the external environment or in the firmsT own programs for expansion or development, and the amount of time that the top management group spent in discussions. The top management groups spent from 42 to 80 percent of their total working time in discussion. In general, the faster the rate of change, the more time the managers spent talking together. Burns' s study also highlighted the amount of lateral communication, particularly in the firms most involved in change, and between those in the lower ranks of management. He found that in each firm there was a senior management group of three, less frequently four, persons, including the general manager, who spent about half their time talking to each other. In many of the companies the managers spent relatively little of their time with their immediate subordinates.




Burns' s study of how these managers spent their time led him to the conclusion that:

The accepted view of management as a working hierarchy on organization chart lines may be dan- gerously misleading. Management simply does not operate as a flow of information up through a succession of filters, and a flow of decisions and instructions down through a succession of ampli- fiers. (13)

More recently, Home and Lupton studied sixty- six middle managers in ten companies which were in different industries in the UK. The companies had between 170 and 40,000 employees. (14) This study also used the self-recording method but it covered only one week, less than Carlson's and Burns' s studies. The headings were similar to those used by Burns with the addition of a classification of all activities into one of four categories: formulating, organizing, unifying, and regulating. Home and Lupton confirmed the findings of Burns and Carlson on the amount of time that managers spend talking. The authors concluded that:

Managers talk most of the time, and mostly face to face. They seem not to be overwhelmed with paper or formal meetings. They swap information and ad- vice and instructions, mostly through informal face to face contact in their own offices. Middle management does not seem, on this showing, to require the exercise of remarkable powers to analyze, weigh al- ternatives, and decide. Rather, it calls for the ability to shape and utilize the person-to-person channels of communication, to influence, to persuade, to facilitate. (15>)

Home and Lupton suggested that on the basis of this evidence, a middle manager needs technical and commercial knowledge of his own firm and an understanding of relationships there.

There have been some other small-scale, self-recording studies, but they have added neither to the methodology nor to our understanding of managers' activities.




The first method of studying what managers do has been used by two American research groups. This is the one where the subjects are asked to estimate how they spend their time. The most ambitious of these studies was that done in the 1950s by the Ohio State University Bureau of Business Research. The Ohio group carried out a series of studies of leadership behavior, mainly among naval officers. Four of these studies sought to determine the effects of different jobs, such as gun- nery officer or dental officer, and of different organizations, upon performance. (16) They used the word "organization" to refer to individual units, such as a mine ship or a landing ship. They distinguished between organizations of different kinds, that is, different kinds of ship or land-based units. They also distinguished between organizations of the same kind, such as two submarines. Information was collected under three headings that were similar to those used in the self-recording studies: time spent in contact with other persons, and who those persons were; time spent in such easily classified activities as observation, writing reports, and preparing charts; time spent in major responsibilities. The last heading contained fourteen categories that included some of those used in the other studies, such as "planning" and "coordination." In addition, four other measures were used. These classified the individual's leadership behavior, not how he spent his time, so they will not be described here.

The Ohio studies are of interest to us for two reasons. One, because the research workers concluded that the type of job was an important factor in determining what its occupant did, more important than the type of organization, though that also had its influence. The other reason is that they were concerned, like the research described in this book, with trying to distinguish different types of jobs on the basis of what their holders did. They distinguished eight job types by analyzing the factors on which individuals in the same job had unusually high or low scores. The job types are not described in detail here as they refer to naval officers1 posts. The titles of the job types were Public Relations Representatives, Professional Consultants,




Personnel Administrators, Technical Supervisors, Schedule- Procedure Makers, Maintenance Administrators, Directors or Decision-Makers, and Coordinators.

A more recent, less ambitious, self-estimating study was carried out by Mahoney, Jerdee, and Carroll at the Manage- ment Development Laboratory of the Industrial Relations Cen- ter at the University of Minnesota, Minneapolis. (1/7) They sought to apply the Ohio State leadership studies to industrial management. They had similar aims: to find out both the similarities and differences in management performance, and how performance varies from one job or organization to another, but they examined fewer aspects of the managers1 activities.

The sample was of 452 managers. They worked for thirteen companies in Minnesota, which were engaged in a wide variety of activities, including manufacturing, finance, maintenance, and public utilities. The number of employees in the companies ranged from 100 to 4,000. Managers at all levels and in all types of functions were included, provided they had one or more subordinates. Nearly all those approached agreed to cooperate, though some were unable to do so. The managers were asked to fill in a questionnaire estimating their usual allocation of time within two classifications of the content of their work. The first had eight headings - planning, investigating, coordinating, evaluating, supervising, staffing, negotiating, and representing. The subjects were also asked to rank these in terms of importance for their job. The second classification had six subject divisions - employees, money and finances, materials and goods, purchases and sales, methods and procedures, and facil- ities and equipment. The authors took great care in their defi- nition of the different headings and in testing the consistency of classifications of specific examples. Managers were given two weeks in which to analyze their performance before completing the check list. The method of analyzing was left to them, so that some managers may have kept quite an accurate diary and others may have done little or no analysis.

Mahoney, Jerdee, and Carroll concluded that an average pro- file of how managers spend their time is misleading; like the




Ohio University group they identified eight job types, but they were different ones. This is not surprising, as the Minnesota group* s criteria for describing the job types were much more limited. They included no information on methods of leadership or on observable activities such as the time spent in contact with other people. The distinction between job types was also based on a much simpler classification; six of the eight job types were identified by 35 to 50 percent of time being spent on one aspect of a manager T s job, such as planning or supervising. The seventh job type was characterized by a relatively high con- centration of time on two aspects which they called multi- specialist. The eighth was called "generalist" because no one or two activities predominated.

The only published study of how executives spend their time that is known to the writer - which used the third method, observation by the research worker - was done by H. Luijk, a Dutch management consultant. (18) Luijk studied twenty-five Dutch directors, each for five days. Most of the recording was done by himself or by one of his secretaries. No analysis was made of the time the directors spent outside their own office except to note the time that they were away. These directors, like the managing directors studied by Carlson, had a large number of visitors and telephone calls. On average each hour was interrupted four times by telephone calls and three times by visitors.

Luijk concluded that on average the directors used about a third of their time inefficiently. (19) He gave nine reasons why he thought that this time had not been well used, which included organization faults and doing too much simple work. He discussed these conclusions with the individuals concerned, who agreed that they could have used that time differently and more profitably. He also thought that twenty-one of his twenty-five subjects allowed their personal preferences to influence their division of time, so that they spent too much time on their fa- vorite subjects and too little on some of the others.

This description of previous studies of how executives spend their time shows that their authors had one or more of four




aims in doing their research. One aim was to describe the average time that managers spent on different activities. This was done by Burns and by Home and Lupton. A second aim was to show up differences in the performance of people in different situations, and to try to find out whether, or how far, these are due to differences in the job, or to differences in the organization. This was the aim of the American studies and, to a lesser extent, of Burns and of Home and Lupton. A third and related aim, which was pursued by the Americans, was to try to distinguish different job types, characterized by the distinctive ways in which their holders spent their time. The last aim, which concerned Carlson and Luijk, was to analyze whether the subjects used their time efficiently. Burns was also interested to find out how his findings compared with the traditional views of how an organization worked. All these aims are relevant to obtaining a better under standing of manager sT jobs and to im- proving the selection and training of managers.

The European studies which have been described used diaries or observation to give us information about what managers actually do. Their methods were more reliable than the two American studies, but they did not attempt to try to distinguish different types of jobs. What is needed next is a study which combines both these approaches. Like the European research, we should study what managers do, as distinct from what they think they do. Like the American research, we should seek to analyze differences between jobs. That is what the study described in this book attempts to do, though it is only a first step because there are so many aspects of managers' jobs that need studying, and only a few that can be included in any one study.

n. JOB PROFILES

The previous chapters [ not included here - Editor] have shown that there were considerable differences in the ways in which the 160 managers spent their time. These variations were so great that it is misleading to talk, as much of the management litera-




ture does, about the managerial job, or about how the average manager spends his time. Yet some differences between managers' jobs are recognized. Thus we talk of junior, middle, or senior managers. We distinguish between works or sales managers, or between staff and line managers. We refer to managers in small, medium, or large companies. Sometimes we talk about managers in a particular industry. Although we refer to these different kinds of managers, we are vague about the ways in which they may differ. We know little about differences in the content of their job or in the ways in which they spend their time. Hence we fall back upon a very general description of what all managers do.

This chapter suggests a new way of classifying managers' jobs. Unlike the distinctions usually made, it is based on an analysis of how managers spend their time. It is only one of the possible kinds of classification since it is limited to the material contained in the diary. Manager s' jobs are so varied and complex that progress towards understanding them requires the gradual building up of a system of classification of the different characteristics of managers' jobs, and of the extent to which jobs vary in their possession of these characteristics.

The data collected in the research was classified by using a special computer program designed by Nigel Howard. This program classifies members of a population, or sample, into groups with common characteristics according to their scores on certain variables. The members of our sample are the 160 managers who kept the diary for four weeks. The twenty-five variables used were derived from the diary entries. [These 25 variables are listed in the Appendix at the end of this article. ] The computer began by dividing the managers into two groups, and then into three, and so on up to ten groups. Each grouping was scrutinized to see if it gave a meaningful and distinctive description of different kinds of managers' jobs. The best one was the division into five groups, but it is worth describing the division into three groups as the division into five is a refine- ment of it. The classification of individual managers into particular groups was compared with the more detailed information




available for each manager to see if the grouping of managers appeared to be a reasonable one. In all cases individuals whose inclusion in a particular group at first seemed odd were found to have special features in their job which explained why they were classified in that way.

In the classification into three groups, Group I consists of the managers who spent much of their time outside the company, traveling, visiting other companies, attending external committees and receiving visitors from outside the company. They worked the longest hours. Group 2 consists of specialist managers who spent much of their time working by themselves: reading, writing, dictating, and calculating. They worked short hours. Group 3 are the average ones in the sample, who spent much of their time talking and listening. Their job was to get other people within their own company to do things. This third group is too large, and contains too many different types of managers, for a three-group classification to be adequate. For this reason Group 3 was subdivided to form two extra groups, making five groups in all. Each group has been given a name, which indicates very broadly one of its main characteristics.

Group 1. The Emissaries

The EmissaryTs work brought him in close touch with the world outside. He spent much of his time away from his company and in talking to people who were not its employees. He often spent several days away, visiting other companies, attending conferences or exhibitions. He worked longer hours than managers in any of the other groups, but this was mainly due to the time spent in traveling and in entertaining. One advantage of his job was that his working day was less fragmented than that of the managers in Groups 3,4, and 5. Since his work was primarily concerned with people outside the company, he was usually less besieged by subordinates and colleagues when he was in his office. He also had more time to himself than members of any other group, except Group 2, though much of this time was unprofitably spent in car, train, or plane.




There are different kinds of Emissaries. The most numerous are those sales managers who had personal contact with customers. Then there are those general managers whose work often took them away from their company, visiting important customers or taking part in professional or trade association activities. Even when they were in their own company, much of their time was spent in contacts with outside people. Group 1 also includes other managers whose work required frequent external contacts, often away from their own company, such as the civil engineers who had to oversee contractors and deal with suppliers. In a larger sample of managers it might, for some purposes, be desirable to subdivide Group 1.

Forty-five of the 160 managers fell into Group 1. These included middle and top managers, in small, medium, and large companies, in a wide variety of industries.

Group 2 . The Writers

This group is markedly different from the others because its members spent more time by themselves in reading, writing, dictating, and figurework. Even so, the Writers were only solitary by comparison with other managers. They spent half their time with other people, compared with two-thirds to three-quar- ters for the average member of the other groups. When they were not working alone they were usually talking with one other person. They spent the least time in group contacts. Group 2 spent less than the average time with every type of contact and less than any of the other groups with colleagues, customers, and other people outside the company.

Members of Group 2 worked shorter hours than the other groups - an average of 39 hours a week. These short hours could not be attributed to less traveling, since they spent slightly more time traveling than members of Groups 3,4, and 5. One reason for the Writers' shorter hours may be that they were more able to control their working day, as they had fewer personal contacts. They were not so much subject to the pressure of day-to-day problems and crises as many of the managers in the other groups.




Group 2 spent about the same amount of time as Groups 3, 4, and 5 within their own establishment, but considerably more in their own office. They rarely had to attend committees, which was one of the reasons why they spent so little time in group contacts. They were more specialized than any of the other groups, though not outstandingly so. Most of their time was spent in their own function.

What kind of managers belonged to this group? It included head office specialists; one would expect to find them spending more time than other managers working on their own. In our sample these were mainly specialist engineering advisers, but some other types of specialist advisers would belong to this group. It also included managers who, although they headed a fair- sized department, were primarily concerned with paperwork, much of which they did on their own. In this category were some of the sales managers who spent their time on the office administration of selling, one payroll manager, and a few accountants and company secretaries. What is surprising - and a salutary reminder of the confusion that can be caused by job titles - is the fact that this group contains several production and works managers whose distribution of time between different activities distinguished them from those of their fellow members of the Institution of Works Managers who are found in Groups 4 and 5.

The line managers who belong to Group 2 differed from its other members in two main ways: they spent more of their time with subordinates and they spent more time on personnel. Both may be explained by the fact that they had more staff than the specialist managers in the group. The line managers in Group 2, including the works and production managers, are distinguished from their colleagues in other groups by the much greater amount of time spent in solitary paperwork. Subordi- nates did not take much less of their time, but they had less contact with people in other departments.

Group 2 includes thirty-three managers in small, medium, and large companies, but the head office specialists are, as one would expect, all in large companies. The other managers




come from companies which vary widely both in size and in type of industry. Most of the group are middle managers, but there are à few top managers and some fairly junior technical specialists.

Group 3. The Discussers

Each of the other four groups had one or more characteristics which distinguished them sharply from the others. This was not true of Group 3, as the proportion of time that its members spent in different activities was close to the average for the whole sample of 160 managers.

Group 3 are called the Discussers because they spent the most time with other people and with their colleagues, but they are not markedly different from Groups 4 and 5 in the amount of their contact time. They spent the same amount of time in contacts involving two or more people as the average for the whole sample, but more time than that of members of the other groups in conversations with one other person. They could be called the "horizontal" group because of the amount of time that they spent with their colleagues, that is, people reporting to the same boss, and because they spent less time than the average with their own staff. Yet they saw more of their boss than the managers in other groups.

What kind of managers are to be found in Group 3 ? The group, which is closest to the average figures for the whole sample, contains, as one might expect, a wider variety of managers than the other groups. It includes all four production engineering managers in the sample, several personnel managers with fair- sized departments, most of the accountants and company secretaries, a number of sales managers who spent most of their time in their own office, one works manager, and the only mer- chandise manager in the sample. Group 3 includes 35 managers from small, medium, and large companies in a variety of industries. It includes middle and senior managers, but no general managers.




Group 4. The Trouble Shooters

Group 4 are the managers with the most fragmented work pattern. This was shown both by the frequency of their diary entries and by their large number of fleeting contacts. This fragmentation arose because they, far more than the managers in the first three groups, had to cope with crises. Even though they may have planned carefully to avoid trouble, much of their time was spent dealing with problems which, when they arose, needed a speedy solution. The repercussions of a failure to solve problems were likely to be more rapid and dramatic than they would be in other departments of the business.

Group 4 could also be called the "man managers," as they spent longer with their subordinates than any other group and longer than any, except Group 5, with the more junior people under their command. Their internal contacts were mainly within the straight-line hierarchy: they spent less time than the sample average with any of the other people in the company. Their main external contacts were with suppliers, with whom they spent more time than any of the other groups.

There is one further way in which Group 4 managers differed from those in other groups, and that is in the relatively large proportion of their time that they spent on inspection. For the other managers this was a small part of their work, but for the Trouble Shooter it occupied more than an hour a day. Most of the members of Group 4 were responsible for a physical area, which they inspected themselves very day.

Most of the works managers come in Group 4. The remainder, with a few exceptions, are in Group 5, along with some production managers. Group 4 contains several factory managers and some who are called works managers, but whose responsibilities are those of a factory manager. There are a few engineering managers who spent a lot of time with their subordinates and on inspection. Group 4 also includes a few general managers of small companies or subsidiaries, who spent nearly all their time in their own company, and who were mainly concerned with works management. The group contains 33 managers from




small, medium, and large companies in a wide variety of industries. It includes works managers in charge of one-off, batch, mass, and process production.

Group 5. The Committeemen

Group 5 differs markedly from the other groups in two re- spects: one, its wide range of internal contacts; and two, the large amount of time its members spent in group discussions. The Committeemen had the widest range of internal contacts because their job, more than that of any of the other groups, involved them in both horizontal and vertical contacts. They had few contacts outside the company. Half their working day was taken up with discussions with more than one person; for most of them this meant that a lot of time was spent in committees. They also differed from other groups in spending more time on personnel work.

Managers in this group worked for large companies. Seven of the fourteen who make up this small group come from the production side of the same company. The seven bridged three levels of production management, from the works manager down- wards. Most of the group are production or works managers. They differed from their counterparts in Group 4 in their less fragmented work pattern, their wider range of internal contacts, and in the much larger proportion of their time that was spent in multiple contacts, usually in committees.

There are three production specialists in the group who had few or no subordinates. They belong to this group because of the wide range of their contacts and the amount of time that they spent in multiple discussions, usually in committees.

The group also includes one accountant whose work brought him into contact with a wide variety of people in his company. He was a departmental head in the accounts department of a large food company. He had 8 people reporting directly to him and 150 in his department. His job had some resemblance to that of a production manager as he was responsible for the day- to-day administration of the mechanized department that pro-




duced all invoices and credit notes and dealt with the related ledger-posting and marketing statistics.

Membership of this group was determined partly by size; only managers who worked for large companies belonged to it. There was some indication that the method of production might also be a factor in determining membership. Nine out of the eleven works and production managers who worked in process industries were in Group 5, and none of the small number who worked in mass production industries. Numbers employed in the works, type of production, and company policies are likely to be three factors that help to determine which group a works manager will belong to.

HI. THE PRESENT SITUATION

Why have there been so few studies in this important area? The explanation may lie, at least in part, in the problems associated with this type of research. There are practical, methodological, and conceptual problems. The main practical problems are obtaining cooperation, and the cost and time of the research. Studying what managers do, as distinct from asking them to fill in questionnaires about their attitudes, is much more de- manding for the manager himself and it is, therefore, harder to find managers who are willing to cooperate. Studying the activities of individual managers, especially if it is done by observation, is costly and time-consuming. The practical difficulties of obtaining a large random sample, using diary-keeping or observation, are probably insuperable.

One of the most important of the methodological problems is how to get a reliable description of managerial activities. The manager's job is a varied and complex one. It is also a very fragmented one; hence, it is very difficult to keep a record of all activities, which has been the aim of many of the studies so far. The method used by some early American studies of asking managers to estimate how they divided their time under different headings has been abandoned in later studies as being too inaccurate. Self-recording diary-keeping is another method,




and one, as described earlier, that has both advantages and limitations. The main limitations are of variations in accuracy between diary-keepers, the omission of many brief activities, and the difficulty of defining subject categories unambiguously. Observation has the advantage of greater accuracy in most re- spects, but it can only be used to study small numbers of managers because of the research time involved, including the vast mass of data that is generated. Two articles in this issue of International Studies of Management & Organization, one by Beishon and Palmer, the other by Wilkie and Young, use different methods of observation. Beishon and Palmer's study is the most detailed and probably the most complete record of the activities of individual managers that is ever likely to be made. One feature of interest in the paper by Graves (also in the present issue of International Studies of Management & Organiza- tion) is that he used a number of different methods: interviews, self-recording diaries, and observation of managers playing a business game.

Another methodological problem for studies which seek to record all a managers' time is that of defining the unit of study. This has usually been called an "episode," and has been defined, for convenience of recording, as lasting as long as there is no change in any of the categories being recorded. This is rather an artificial way of distinguishing different activities. Its arti- ficiality may not matter where all activities are summed, but can lead to differences between studies if they adopt different definitions of an episode. The main difference, as in the Beishon and Palmer and the Wilkie and Young studies, is likely to be in the number of episodes recorded.

The principal theoretical problem of this kind of research, although there has been little discussion of it in the literature, is that of the conceptualization of the manager1 s job. There is the traditional concept of management functions, such as planning, organizing, and controlling. This is the only one that has gained wide acceptance, but it is of very limited value in describing managerial behavior. It tells one little about the manager's job and nothing about the differences between jobs. Most




of the studies of managerial activities so far have used few or no explicit concepts about managers' jobs. Instead they have sought, using episodes as the unit of classification, to record everything that the manager does. Obviously, any type of recording requires a classification, which in turn requires decisions about what is to be classified and how. Some of the attempts made in early studies to classify the content of man- agers1 work suffered from being noncomparable because of the ambiguity and vagueness of the terms used, like "planning."

The studies so far have managed to contribute to our understanding of managerial behavior and the nature of managers' jobs by using very simple categories, such as where the manager is working, whether he is alone or with somebody else, who he is talking to, and simple classifications of methods of work such as writing, attending meetings, telephoning, and an analysis of the frequency of activities. The latter is the source of one of the main contributions of these studies to date. They have highlighted - and Beishon and Palmer T s study does this most of all - how very fragmented the manager's work is. Probably there is not much more to be contributed in further research by the use of these restricted and simple categories for looking at a manager's activities; nor by the addition, as in some sutdies, of ambiguous functional categories.

We now need better concepts of the nature of managerial work, so as to develop more sophisticated forms of classification of managers' activities. One way of doing so, which is represented in this issue by the articles of Graves and Heller, is to concen- trate on a particular aspect of managerial behavior, where there are concepts that have already been developed and can be used. Another and promising approach is that used by Mintzberg (20) who has classified managers' activities into ten different roles, which are aspects of the three basic behaviors that he identified: interpersonal, information processing, and decision-making. Further work will be needed by other research workers to see if they find Mintzberg' s categories useful for observing other managers. But his approach offers a promising depar- ture from the traditional classification of management functions.




Two further research needs stand out from this account of the problems and approaches of studies of managerial behavior so far. The first is to seek to develop our conceptualization of the manager's job. The second is to study the differences as well as the similarities in managerial behavior and in managers1 jobs.

APPENDIX

List of the Twenty-Five Variables Used for the Classification into Job Types

Each manager was classified on the basis of his score for the four weeks on each of the variables. The variables were given equal weight. 1. Total hours worked 2. Total number of entries. 3. Total number of fleeting contacts. 4. Percentage of total time spent alone. 5. " " with one other person. 6. M " with two or more other people. 7. " M in his own establishment (not including

time spent in other units of the company). 8. " " outside the company. 9. " " in traveling.

10. " " with his boss. 11. " " with his secretary. 12. " " with his immediate subordinates. 13. " " with his subordinatesT subordinates. 14. " " with his colleagues. 15. " " with his fellow specialists. 16. " " with other internal contacts. 17. " " with customers. 18. " " with suppliers. 19. M M with other external contacts. 20. " n in committees. 21. TT "in all other forms of discussion.




22. Percentage of total time in all forms of paperwork. 23. " " in inspection. 24. " " spent on personnel matters. 25. " TT spent on the four functional areas

which took least time. (This was included as a guide to the range of management functions with which the manager was concerned.)

Notes

1) Rosemary Stewart, The Reality of Management, London, Heinemann, 1963.

2) Elliot D. Chappie and Leonard R. Sayles, The Measure of Management: Designing Organizations For Human Effective- ness, New York, Macmillan, 1961.

3) Joan Doulton and David Hay, Managerial and Professional Staff Grading, London, Allen & Unwin, 1962.

4) Elliott Jaques, Measurement of Responsibility, London, Tavistock Publications, 1957.

5) Elliott Jaques, Equitable Payment: A General Theory of Work, Differential Payment, and Individual Progress, London, Heinemann, 1961, p. 17. The previous reference was to Jaques* s first book on comparing managers' jobs.

6) "Contacts" is used for the people that a manager talks with in the course of his work.

7) Burns, in his study of 76 managers, found that there was a general tendency to overestimate the time taken by the main divisions of management such as production and accounts and to underestimate the time taken by personnel matters and discussions on general policy. See T. Burns, "Management in Ac- tion," Operational Research Quarterly, Vol. VIH, No. 2 (June 1957), pp. 45-60.

Mahoney, Jerdee, and Carroll, who did a study of estimated distribution of time, which is discussed later, checked the reliability of the estimates by work- sampling four managers. They said that only approximately 56 percent of assignments




were classified in the same way by both methods. See T. A. Mahoney, J. H. Jerdee, and S. J. Carroll, Development of Man- agerial Performance ... A Research Approach, Monograph C-9, South-Western Publishing, USA, 1963.

Carlson in his study found that the managing directors had little idea of how they spent their time. See Sune Carlson, Ex- ecutive Behavior: A Study of the Workload and Working Meth- ods of Managing Directors, Strömbergs, Stockholm, 1951.

Another check of the reliability of estimates found that there was a much closer correlation between estimated and recorded time under some headings than under others. There was a fairly high correspondence for such activities as writing reports, reading mail, talking with other people, operating machines and attending meetings, but no correlation at all for planning. See Ralph M. Stogdill, Shartle, Carroll, et al., Patterns of Adminis- trative Performance, Ohio State University, Bureau of Busi- ness Research, Monograph No. 81 (1956), p. 48.

8) A random alarm device, about the size of a packet of cigarettes, was used in a study of the reading behavior of chem- ists and physicists. The Operations Research Group of the Case Institute of Technology in Cleveland, Ohio, studied 701 scientists for fourteen days. The random alarm went off on average 3.5 times a day. The research is described in Miles W. Martin Jr., "The Use of Random Alarm Devices in Studying Scientists' Reading Behaviour," IRE Transactions of the Pro- fessional Group of Engineering Management, Vol. EM-9, No. 2 (June 1962).

9) Carlson, op. cit. 10) The word "function" is used in many different ways by

writers on management. It is used in this book to refer to the main activities of the business such as sales, production, finance, and research and development. An alternative word might have been "department" except that one function may be composed of many departments.

11) Carlson, op. cit., p. 114. 12) Burns, op. cit. 13) Ibid., p. 60.




14) J. H. Home and Tom Lupton, "The Work Activities of 'Middle T Managers/1 The Journal of Management Studies, Vol. I, No. 2 (Feb. 1965), pp. 14-33.

15) Ibid., p. 32. 16) Stogdill, Shartle, Carroll, et al., op. cit. 17) Mahoney, Jerdee, and Carroll, op. cit. 18) G. Copeman, H. Luijk, and F. de P. Hanika, How the Ex-

ecutive Spends His Time, Business Publications, London, 1963. Thurley and Hamblin did an observation study of supervisors

which used the incident method. See K. E. Thurley and A. C. Hamblin, "The Supervisor and His Job,M Problems of Progress in Industry, No. 13. Department of Scientific and Industrial Re- search (London, H.M.S.O., 1963).

19) How the Executive Spends His Time, p. 66. 20) Henry Mintzberg, "Managerial Work: Analysis from Ob-

servation," Management Science, Vol. 18, No. 2 (October 1971), pp. 97-110.



the study of managerial behavior || studies of managerial jobs: methodologies and profiles

Documents