Rounding out the Manager’s Job

Reading Time: 47 min 

Tom Peters tells us that good managers are doers. (Wall Street says they “do deals.”) Michael Porter suggests that they are thinkers. Not so, argue Abraham Zaleznik and Warren Bennis: good managers are really leaders. Yet, for the better part of this century, the classical writers — Henri Fayol and Lyndell Urwick, among others — kept telling us that good managers are essentially controllers.

It is a curiosity of the management literature that its best-known writers all seem to emphasize one particular part of the manager’s job to the exclusion of the others. Together, perhaps, they cover all the parts, but even that does not describe the whole job of managing.

If you turn to the more formalized literature, you will find all kinds of lists — of tasks or roles or “competences.” But a list is not a model (even if presented in the form of a circle, meaning the ends have been joined), and so the integrated work of managing still gets lost in the process of describing it. And without such a model, we can neither understand the job properly nor deal with its many important needs — for design, selection, training, and support.

To play with a metaphor, if the toughest nut to crack in our knowledge of management has been the manager’s job itself, then that may well be because we have done just that. We have been so intent on breaking the job into pieces that we never came to grips with the whole thing. It is time, therefore, to consider the integrated job of managing.

That is what I set out to do several years ago, after becoming discouraged with all those lists and circles (including one from my own initial study of managerial work, first published in 1973).1 I did not feel the need to go find out what managers do. We knew that already, I believed, based on a considerable body of research and publication over the past decades. Our need was for a framework to put all this together, a model of managing, if you like. People had to be able to “see” the job in one place, in order to deal with its component parts comprehensively and interactively. In fact, as my ideas developed, the metaphor of the nut came alive, for the model has taken the form of an interacting set of concentric circles.

Using this model, I began to spend time with managers at work, both to check out and to flesh out the model, especially as it applies to different managerial jobs and styles. I have been spending a day with each of a number of managers — observation reinforced by interviewing — not to draw any definitive conclusions so much as to get a flavor for as wide a variety of managerial jobs and styles as possible. So far they number twenty-three, ranging from the head of one of Europe’s largest state health care systems to the “front country” manager of a Canadian mountain park. (Examples used throughout this paper will give a sense of the variety of managers I have studied so far.)

This article presents the model, building the image of the manager’s job from the inside out, beginning at the center with the person and his or her frame and working out from there, layer by layer. Once this description is complete, I shall discuss briefly the jobs and styles of some of the managers I have observed to give a sense of how readers might use the model to better appreciate their own jobs or those of the managers around them. I shall also comment on the effective practice of managerial work, concluding, as you may have already guessed, that this is one job that has to be “well rounded”!

The Person in the Job

We begin at the center, with the person who comes to the job. People are not neutral when they take on a new managerial job, mere putty to be molded into the required shape. Indeed, greater appreciation of this fact would allow us to be more careful in how we select managers in the first place. Or else more flexible in how we let them mold their jobs to themselves in the second.

Figure 1 shows that an individual comes to a managerial job with a set of values, by this stage in life probably rather firmly set — for example, that the radio station he manages should provide intelligent fare. He or she also brings a body of experience that, on one hand, has forged a set of skills or competences, perhaps honed by training, and, on the other, has provided a base of knowledge — that comes, for example, from spending thirty-five years on a major police force before becoming its commissioner. That knowledge is, of course, used directly, but it is also converted into a set of mental models, key means by which managers interpret the world around them — for example, how the head nurse on a hospital ward perceives the behavior of the surgeons with whom she must work. Together, all these characteristics greatly determine how any manager approaches a given job — his or her style of managing. Style will come to life as we begin to see how a manager carries out what his or her job requires.

The Frame of the Job

Embed the person depicted in a given managerial job and you get managerial work. At the core of it is some kind of frame for the job, the mental set the incumbent assumes to carry it out. Frame is strategy, to be sure, possibly even vision, but it is more than that. We can show it as three increasingly specific components, emanating from the person concentrically in Figure 2.

First is purpose, namely what the manager is seeking to do fundamentally with the unit he or she is supposed to manage — for example, to increase the state funding for a hospital or to open more stores in a retail chain. As depicted around the circle, a manager may create a unit, maintain the effective functioning of a unit already created, adapt the unit to some new conditions, or recreate in some more ambitious way a unit previously created. To use a metaphor, a manager may build the track in the first place, keep the unit on track (or put it back on track), improve or shift the track, or else rebuild the track. The head nurse seeks to keep her unit running smoothly; the chief executive of the state health care system accepts a government mandate for major restructuring.

The two circles beyond purpose describe two further dimensions of frame, encompassing both strategy and structure. First is perspective, equivalent to what Peter Drucker has called a “theory of the business” (or of any unit being managed).2 Perspective is the overall approach to the management of the unit, including such notions as “vision” and “culture.” Second are positions, which, in contrast, are more concrete, closer to Michael Porter’s view of strategy and perhaps many consulting companies’ view of structure. These positions consider specific locations for the unit in its environment, and specific ways of doing its work, such as the products produced, the markets served, the structures and systems designed, the facilities provided.3

Alain Noël, who studied the relationship between the frames and the work of the chief executives of three small companies, has said that managers have “occupations” and they have “preoccupations.”4 Frame describes the preoccupations, while roles (discussed later) describe the occupations. But frame does give rise to a first role in this model as well, which I call conceiving, namely thinking through the purpose, perspective, and positions of a particular unit to be managed over a particular period of time.

Different managers conceive their frames in different ways. In other words, the style of performing this first role can vary significantly. First, the frame can be imposed by some outside person or force, or else it can be developed by the manager him- or herself. And second, that frame may range from being rather sharp to rather vague, for example, from “cutting costs by 10 percent before year end” to “getting this place in order.” The frame of any managerial job, as suggested by its placement in the center of the rounding model, is a kind of magnet for the behaviors that surround it. So long as the frame is rather sharp, it holds those behaviors together tightly. But when the frame is vague, the different issues considered and the different activities performed risk flying off in all directions. That is presumably why there has been so much demand in recent years for clearer “vision” in strategic thinking.

As shown in Table 1, these two sets of dimensions produce four broad styles of conceiving the frame. A frame that is self-selected but vague allows the manager wide latitude to maneuver but offers little real sense of direction. Managerial style is likely to become opportunistic, as in the case of a health care manager who uses any possible means to bring his unit under control. But having a vague frame imposed on the manager (such as “empower your people”) may provide little real help and risks evoking a passive style of management (unless, of course, the manager sharpens it). A sharp frame that the manager selects would tend to lead to a determined, sometimes “visionary,” style of management. For example, the head of a fashion museum sees her role as the preservation of a national heritage. A sharp frame that is, in contrast, imposed, could well lead to a driven style of management, as in the example of that chief executive of the state medical system who tries to honor the government’s intentions.

The Agenda of the Work

Given a person in a particular managerial job with a particular frame, the question arises of how this is manifested in the form of specific activities. That happens through the agenda to carry out the work, and the associated role of scheduling, which has received considerable attention in the literature of management, for example, in the empirical work of John Kotter.5 Agenda is considered in two respects here, again shown as concentric circles (in Figure 3), the inner one more general, the outer one more specific.

First, the frame gets manifested as a set of current issues, in effect, whatever is of concern to the manager, broken down into manageable units — what Tom Peters likes to call “chunks.” Ask any manager about his or her work, and the almost inevitable first reply will be about the “issues” of central concern, those things “on the plate,” as the saying goes — promoting public exhibitions at the museum and developing a state-of-the-art storage facility for its clothing, for example. Or take a look at the agendas of meetings and you will likewise see a list of issues (rather than decisions). These, in effect, operationalize the frame (as well as change it, of course, by feeding in new concerns).

As already noted, the sharper the frame, the more integrated the issues. The more realizable they may be as well, since it is a vague frame that gives rise to that all-too-common phenomenon of the unattainable “wish list” in an organization. Sometimes a frame can be so sharp, and the issues therefore so tightly integrated, that they all reduce to what Noël has called one “magnificent obsession.”6 In effect, all the concerns of the manager revolve around one central issue, for example, imposing those changes on the state health care system or expanding that retail chain.

Second, the frame and the issues get manifested in the more tangible schedule, the specific allocations of managerial time on a day-by-day basis. Also included here, however implicitly, is the setting of priorities among the issues. The scheduling of time and the prioritization of issues are obviously of great concern to all managers, and, in fact, are themselves significant consumers of managerial time. Accordingly, a great deal of attention has been devoted to these concerns, including numerous courses on “time management.”

The Core in Context

If we label the person in the job with a frame manifested by an agenda (i.e., all the circles so far discussed), the central core of the manager’s job, then we turn next to the context in which this core is embedded, the milieu in which the work is practiced.

The context of the job is depicted in Figure 4 by the lines that surround the core. I have so far been using the word “unit” rather freely. Let me be more specific. A manager, by definition, has formal authority over an organizational unit, whether that be a whole organization in the case of a chief executive or a division, department, or branch, and so on, in the case of a manager within the hierarchy. Context can thus be split into three areas, labeled inside, within, and outside on Figure 4.

Inside refers to the unit being managed, shown below the manager to represent his or her formal authority over its people and activities — the hospital ward in the case of the head nurse, for example. Within, shown to the right, refers to the rest of the organization, other members and other units with which the manager must work but over which he or she has no formal authority — the doctors, the kitchen, the physiotherapists in the rest of the hospital, to continue with the same example. (Of course, in the case of chief executive, there is no inside separate from within: that person has authority over the entire organization.) And outside refers to the rest of the context not formally part of the organization with which the manager must work — in this example, patients’ relatives, long-term care institutions to which some of the unit’s patients are discharged, nursing associations, and so on. The importance of this distinction (for convenience, we shall mostly refer to inside versus outside) is that much of managerial work is clearly directed either to the unit itself, for which the manager has official responsibility, or at its various boundary contexts, through which the manager must act without that responsibility.

Managing on Three Levels

We are now ready to address the actual behaviors that managers engage in to do their jobs. In other words, we turn now from the largely cerebral roles of conceiving and scheduling at the core to the more tangible roles of getting things done. The essence of the model, designed to enable us to “see” managerial work comprehensively, is that these roles are carried out on three successive levels, each inside and outside the unit. Again, this is depicted by concentric circles of increasing specificity, shown in Figure 5.

From the outside (or most tangible level) in, managers can manage action directly, they can manage people to encourage them to take the necessary actions, and they can manage information to influence the people in turn to take their necessary actions. In other words, the ultimate objective of managerial work, and of the functioning of any organizational unit, the taking of action, can be managed directly, indirectly through people, or even more indirectly by information through people. The manager can thus choose to intervene at any of the three levels, but once done, he or she must work through the remaining ones. Later we shall see that the level a given manager favors becomes an important determinant of his or her managerial style, especially distinguishing so-called “doers” who prefer direct action, “leaders” who prefer working through people, and “administrators” who prefer to work by information.

The discussion that follows describes managerial behavior in terms of roles, each pertaining to a given level and directed either inside or outside the unit (or both). We shall begin with the innermost circle, of information, containing the most conceptual roles, and work out to the more tangible or specific ones.

In presenting these roles, it should be emphasized that all managers perform all of them as the essence of their work. Styles of managers do vary, to be sure, but not in whether these roles are performed so much as in which of them is favored and how they are performed. This model of roles on levels directed inside and outside the unit thus provides the basis for understanding different styles of managing as well as different contexts in which that managing takes place.

Managing by Information

To manage by information is to sit two steps removed from the purpose of managerial work. The manager processes information to drive other people who, in turn, are supposed to ensure that necessary actions are taken. In other words, here the managers’ own activities focus neither on people nor on actions per se, but rather on information as an indirect way to make things happen. Ironically, while this was the classic perception of managerial work for the first half of this century (as I discuss later), in recent years, it has also become a newly popular, in some quarters almost obsessional, view, epitomized by the so-called “bottom line” approach to management.

The manager’s various informational behaviors may be grouped into two broad roles, here labeled communicating and controlling, shown in Figure 6.

Communicating refers to the collection and dissemination of information. In Figure 6, communicating is shown by double arrows to indicate that managers devote a great deal of effort to the two-way flow of information with the people all around them — employees inside their own units, others in the rest of the organization, and especially, as the empirical evidence makes abundantly clear, a great number of outsiders with whom they maintain regular contact. Thus the head of one regional division of the national police force spent a good part of his day passing information back and forth between the central headquarters and the people on his staff.

Managers “scan” their environments, they monitor their own units, and they share with and disseminate to others considerable amounts of the information they pick up. A point worth emphasizing, and one emphasized in almost every serious study of managerial work, is that the formal information — in other words, information capable of being processed in a computer — does not play a particularly dominant role here. Oral information — much of it too early or too “soft” to formalize, such as gossip and hearsay — and even nonverbal information, namely what is seen and “felt” but not heard, forms a critical part of every serious managerial job (or, at least, every managerial job performed seriously).

In my initial study, I described managers as “nerve centers” of their units, who use their status of office to gain access to a wide variety of informational sources. Inside the unit, everyone else is a specialist who generally knows more about his or her specialty than the manager. But, because the manager is connected to all those specialists, he or she should have the broadest base of knowledge about the unit in general. This should apply to the head of an 800,000-person health care system, with regard to broad policy issues, no less than to the clinical director of one of its hospital units, with regard to the service rendered there. And externally, by virtue of their status, managers have access to other managers who are themselves nerve centers of their own units. And so they tend to be exposed to powerful sources of external information and thus emerge as external nerve centers as well. The health care chief executive can thus talk to people running health care systems in other countries and so gain access to an array of information perhaps inaccessible even to his most influential reports.

The result of all this is that a considerable amount of the manager’s information turns out to be privileged, especially when we consider how much of it is oral and nonverbal. Accordingly, to function effectively with the people around them, managers have to spend considerable time sharing their information, both with outsiders (in a kind of spokesperson role) and with insiders (in a kind of disseminator role).

I found in my initial study of chief executives that perhaps 40 percent of their time was devoted almost exclusively to the communicating role — just to gaining and sharing information — leaving aside the information processing aspects of all the other roles. In other words, the job of managing is fundamentally one of processing information, notably by talking and especially listening. Thus Figure 6 shows the inner core (the person in the job, conceiving and scheduling) connected to the outer rings (the more tangible roles of managing people and action) through what can be called the membrane of information processing all around the job.

What can be called the controlling role describes the managers’ efforts, not just to gain and share information, but to use it in a directive way inside their units: to evoke or provoke general action by the people who report to them. They do this in three broad ways: they develop systems, they design structures, and they impose directives. Each of these seeks to control how other people work, especially with regard to the allocation of resources, and so what actions they are inclined to take.

First, developing systems is the most general of these three, and the closest to conceiving. It uses information to control people’s behaviors. Managers often take charge of establishing and even running such systems in their units, including those of planning and performance control (such as budgeting). The head nurse, for example, dispensed with one of the hospital’s key control systems and instead developed her own. Robert Simons has noted how chief executives tend to select one such system and make it key to their exercise of control, in a manner he calls “interactive.”7

Second, managers exercise control through designing the structures of their units. By establishing responsibilities and defining hierarchial authority, they again exercise control rather passively, through the processing of information. People are informed of their duties, which in turn is expected to drive them to carry out the appropriate actions. Thus, a day spent with the head of an international environmental organization involved considerable attention to reorganization.

Third is imposing directives, which is the most direct of the three, closest to the people and action, although still informational in nature. Managers pronounce: they make specific choices and give specific orders, usually in the process of “delegating” particular responsibilities and “authorizing” particular requests. In effect, managers manage by transmitting information to people so that they can act. Thus a deputy minister in the Canadian government met with a number of his policy analysis people to provide rather specific comments about his wishes on drafts of reports they had submitted to him.

If a full decision-making process can be considered in the three stages of diagnosing, designing, and deciding — in other words, identifying issues, working out possible solutions, and selecting one8 — then here we are dealing with a restricted view of decision making. Delegating means mostly diagnosing (“Would you please handle this problem in this context”), while authorizing means mostly deciding (“OK, you can proceed”). Either way, the richest part of the process, the stage of designing possible solutions, resides with the person being controlled rather than with the manager him- or herself, whose own behavior remains rather passive. Thus the manager as controller seems less an actor with sleeves rolled up, digging in, than a reviewer who sits back in the office and passes judgment. That is why this role is characterized as informational; I will describe a richer approach to decision making in the section on action roles.

The controlling role is shown in Figure 6 propelling down into the manager’s own unit, since that is where formal authority is exercised. The single-headed arrows represent the imposed directives, while the pitchfork shape symbolizes both the design of structure and the development of systems. The proximity of the controlling role in Figure 6 to the manager’s agenda reflects the fact that informational control is the most direct way to operationalize the agenda, for example, by using budgets to impose priorities or delegation to assign responsibilities.

The controlling role is, of course, what people have in mind when they refer to the “administrative” aspect of managerial work. Interestingly, it encompasses almost the entire set of activities described by the classical writers. In the 1930s, for example, Gulick and Urwick popularized the acronym POSDCORB (planning, organizing, staffing, directing, coordinating, reporting, and budgeting).9 Planning, organizing, directing, and budgeting are all clearly focused here, while reporting, coordinating, and staffing have important, although not exclusive, controlling aspects (staffing in the sense of deciding). Thus it must be concluded that the long-popular description of managerial work was not so much wrong as narrow, focusing almost exclusively on one restricted aspect of the job: informational control of the unit through the exercise of formal authority.

Managing through People

To manage through people, instead of by information, is to move one step closer to action, but still to remain removed from it. That is because here the focus of managerial attention becomes affect instead of effect. Other people become the means to get things done, not the manager him- or herself, or even the substance of the manager’s thoughts.

After several decades of POSDCORB thinking and Taylorist technique, the Hawthorne experiments of the 1930s demonstrated with dramatic impact that management has to do with more than just the passive informational control of subordinates.10 People entered the scene, or at least they entered the textbooks, as entities to be “motivated” and later “empowered.” Influencing began to replace informing, and commitment began to vie with calculation for the attention of the manager. Indeed, in the 1960s and 1970s especially, the management of people, quite independent of content — of the strategies to be realized, the information to be processed, even the actions to be taken — became a virtual obsession of the literature, whether by the label of “human relations,” “Theory Y,” or “participative management” (and later “quality of work life,” to be replaced by “total quality management”).

For a long time, however, these people remained “subordinates” in more ways than one. “Participation” kept them subordinate, for this was always considered to be granted at the behest of the managers still fully in control (compared with the more constitutional involvement of certain professionals, such as doctors in hospitals, or even of certain European workers who, under “co-determination,” gained legal representation on boards of directors). So does the currently popular term “empowerment,” which implies that power is being granted, thanks to the managers. (Hospital directors do not “empower” physicians!) People also remained subordinates because the whole focus was on those inside the unit, not outside it. Not until serious research on managerial work began did it become evident how important to managers were contacts with individuals outside their units (who, for example, go entirely without mention in POSDCORB!). Virtually every single study of how all kinds of managers spent their time has indicated that outsiders, of an enormously wide variety, generally take as much of managers’ attention as so-called “subordinates.” We shall thus describe two people roles here, shown in Figure 7, one internal, called leading, and one external, called linking.

The leading role has probably received more attention in the literature of management than all the other roles combined. And so we need not dwell on it here. But neither can we ignore it: managers certainly do much more than lead the people in their own units, and leading certainly infuses much else of what managers do (as, in fact, do all the roles, as we have already noted about communicating). But their work just as certainly cannot be understood without this dimension.11 We can describe the role of leading on three levels, as indicated in Figure 7.

First, managers lead on the individual level, “one on one,” as the expression goes. They encourage and drive the people of their units — motivate them, inspire them, coach them, nurture them, push them, mentor them, and so on. All managers, from the chief executive of the major police force to the front-country manager in the mountain park, stop to chat with their people informally during the day to encourage them in their work. Second, managers lead on the group level, especially by building and managing teams, an effort that has received considerable attention in recent years. Again, team meetings, including team building, figured in many of my observations, for example, the head of a London film company who brought film-making teams together for both effective and affective purposes. And third, they lead on the unit level, especially with regard to the creation and maintenance of culture, another subject of increasing attention in recent years (thanks especially to the Japanese). Managers, for example, engage in many acts of a symbolic nature (“figurehead” duties) to sustain culture, as when the head of the national police force visited its officer training institute (as he does frequently) to imbue the force’s norms and attitudes in its graduating class.

All managers seem to spend time on all three levels of leadership, although, again, styles do vary according to context and personality. If the communicating role describes the manager as the nerve center of the unit, then the leading role must characterize him or her as its “energy center,” a concept perhaps best captured in Maeterlinck’s wonderful description of the “spirit of the hive.”12 Given the right managerial “chemistry” (in the case of Maeterlinck’s queen bee, quite literally!), it may be the manager’s mere presence that somehow draws things together. By exuding that mystical substance, the leader unites his or her people, galvanizing them into action to accomplish the unit’s mission and adapt it to a changing world.

The excess attention to the role of leading has probably been matched by the inadequate attention to the role of linking. For, as already noted, in their sheer allocation of time, managers have been shown to be external linkers as much as they are internal leaders. In 1964, Leonard Sayles emphasized this in his path-breaking book, and I repeated it in 1973, as did John Kotter in 1982.13 Yet, still the point seems hardly appreciated. Indeed, now more than ever, it must be understood, given the great growth of joint ventures and other collaborating and networking relationships between organizations, as well as the gradual reconception of the “captive” employee as an autonomous “agent” who supplies labor.

Figure 7 suggests a small model of the linking role. The arrows go in and out to indicate that the manager is both an advocate of its influence outside the unit and, in turn, a recipient of much of the influence exerted on it from the outside. In the middle are two parallel lines to represent the buffering aspect of this role — that managers must regulate the receipt of external influence to protect their units. To use a popular term, they are the “gatekeepers” of influence. Or, to add a metaphor, the manager acts as a kind of valve between the unit and its environment. Nowhere was this clearer than in my observation of three levels of management in a national park system — a regional director, the head of one mountain park, and the front-country manager of that park. They sit in an immensely complex array of forces — developers who want to enhance their business opportunities, environmentalists who want to preserve the natural habitat, tourists who want to enjoy the beauty, truckers who want to drive through the park unimpeded, politicians who want to avoid negative publicity, etc. It is a delicate balancing, or buffering, act indeed!

All managers, as emphasized in other research and my studies cited earlier, appear to spend a great deal of time “networking” — building vast arrays of contacts and intricate coalitions of supporters beyond their own units, whether within the rest of the organization or outside, in the world at large. To all these contacts, the manager represents the unit externally, promotes its needs, and lobbies for its causes. In response, these people are expected to provide a steady inflow of information to the unit as well as various means of support and specific favors for it. This networking was most evident in the case of the film company managing director, who even in one day exhibited an impressive network of contacts in order to negotiate her complex contracts with various media in different countries.

In turn, people intent on influencing the behavior of an organization or one of its subunits will often exercise pressure directly on its manager, expecting that person to transmit the influence inside, as was most pointedly clear in the work of the parks managers. Here, then, the managerial job becomes one of delicate balance, a tricky act of mediation. Those managers who let external influence pass inside too freely —who act like sieves — are apt to drive their people crazy. (Of course, those who act like sponges and absorb all the influence personally are apt to drive themselves crazy!) And those who block out all influence —who act like lead to x-rays — are apt to detach their units from reality (and so dry up the sources of external support). Thus, what influence to pass on and how, bearing in mind the quid pro quo that influence exerted out is likely to be mirrored by influence coming back in, becomes another key aspect of managerial style, worthy of greatly increased attention in both the study of the job and the training of its occupants.

Managing Action

If managers manage passively by information and affectively through people, then they also manage actively and instrumentally by their own direct involvement in action. Indeed, this has been a long-established view of managerial work, although the excess attention in this century, first to controlling and then to leading, and more recently to conceiving (of planned strategy), has obscured its importance. Leonard Sayles, however, has long and steadily insisted on this, beginning with his 1964 book and culminating in The Working Leader (published in 1993), in which he makes his strongest statement yet, insisting that managers must be the focal points for action in and by their units.14 Their direct involvement must, in his view, take precedence over the pulling force of leadership and the pushing force of controllership.

I shall refer to this involvement as the doing role. But, in using this label — a popular one in the managerial vernacular (“Mary Ann’s a doer!”) — it is necessary to point out that managers, in fact, hardly ever “do” anything. Many barely even dial their own telephones! As already noted, watch a manager and you will see someone whose work consists almost exclusively of talking and listening, alongside, of course, watching and “feeling.” (That, incidentally, is why I show the manager at the core of the model as a head and not a full body!)

What “doing” presumably means, therefore, is getting closer to the action, ultimately being just one step removed from it. Managers as doers manage the carrying out of action directly, instead of indirectly through managing people or by processing information. In effect, a “doer” is really someone who gets it done (or, as the French put it with their expression faire faire, to make something get made). And the managerial vernacular is, in fact, full of expressions that reflect just this: “doing deals,” “championing change,” “fighting fires,” “juggling projects.” In the terms of decision making introduced earlier, here the manager diagnoses and designs as well as decides: he or she gets deeply and fully involved in the management of particular activities. Thus, in the day I spent with the head of the small retail chain, I saw a steady stream of all sorts of people coming and going, most involved with some aspect of store development or store operations, and there to get specific instructions on how to proceed next. He was not delegating or authorizing, but very clearly managing specific developmental projects step by step.

Just as they communicate all around the circle, so too do managers “do” all around it, as shown in Figure 8. They manage projects and solve problems, or put out fires, inside their units, and they “do deals” and negotiate agreements with outsiders. Let us consider each in turn.

Doing inside involves projects and problems. This does not imply that projects are independent of problems, or that either are exclusively internal. Rather, it means that much “doing” has to do with changing the unit itself, both proactively and reactively. Managers champion change to exploit opportunities for their units, and they handle its problems and resolve its crises, often with “hands on” involvement. Indeed, the president of a large French systems company spent part of his day in a meeting on a very specific customer contract. Asked why he attended, he said it was a leading-edge project that could well change his company. He was being informed, to be sure, but also “doing” (more than controlling): he was an active member of the team.

The difference between deciding in the controlling role and acting in the doing role is the difference, in effect, between sitting in a managerial office and passing judgment on the issues floating by — “Roberta, please handle this,” “No, Joe, you can’t do that” — and getting out of the office and actively engaging an issue from its initial identification to its final resolution. Here the manager becomes a true designer (or, in the example above, a partner in the design), not of abstract strategies or of generalized structures, but of tangible projects of change. And the evidence, in fact, is that managers at all levels typically juggle many such projects concurrently, perhaps several dozen in the case of chief executives. Hence the popularity of the term “project management.”

Managers “do” inside in two other respects as well. For one thing, they substitute, sometimes doing the routine work of their units in place of other people. (The front-country manager nearly spent part of the day cruising a river for the body of a person who had a few days earlier gone over a waterfall in a boat!) Of course, when managers are replacing absent employees, this may be considered another aspect of handling problems. Second, some managers continue to do regular work after they have become managers. For example, the head nurse saw a patient (just as the Pope leads prayers, or a dean might teach a class). Done for its own sake, this might be considered separate from managerial work. But such things are often done for very managerial reasons as well. This may be an effective way of “keeping in touch” with the unit’s work and finding out about its problems, in which case it falls under the role of communicating. Or it may be done to demonstrate involvement and commitment with others in the unit, in which case it falls under the role of culture building in the role of leading.

Doing outside takes place in terms of deals and negotiations. Again, these are two sides of the same coin, in that managers negotiate in order to “do deals,” and also negotiate once the deal is done. And, again as well, there is no shortage of evidence on the importance of negotiating as well as dealing in managerial work. Most evident in my observations was the managing director of the film company, who was working on one intricate deal after another that day. This was a small company, and making deals were a key part of her job. In larger organizations, senior managers may have all kinds of specialized negotiators supporting them (for example, lawyers for contracts and labor relations specialists for union bargaining). Yet that does not release them from having to spend considerable time on negotiations themselves, especially when critical moments arise. After all, they are the ones who have the authority to commit the resources of their unit, and it is they who are the nerve centers of its information as well as the energy centers of its activity, not to mention the conceptual centers of its strategy. All around the circles, therefore, action connects to people who connect to information, which connects to the frame.

The Well-Rounded Job of Managing

I opened this article by noting that the best-known writers of management all seem to emphasize one aspect of the job — in the terms we now have, “doing” for Tom Peters, “conceiving” for Michael Porter, “leading” for Abraham Zaleznik and Warren Bennis, “controlling” for the classical writers, and so on. Now it can be appreciated why all may well be wrong: heeding the advice of any one of them must lead to the lopsided practice of managerial work. Like an unbalanced wheel at resonant frequency, the job risks flying out of control. That is why it is important to show all of the components of managerial work on a single integrated diagram, as in Figure 9, to remind people, at a glance, that these components form one job and cannot be separated.

Acceptance of Tom Peters’ urgings — “ ‘Don’t think, do’ is the phrase I favor”15 — could lead to the centrifugal explosion of the job, as it flies off in all directions, free of a strong frame anchoring it at the core. But acceptance of the spirit of Michael Porter’s opposite writings —that what matters most is conception of the frame, especially of strategic positions — could produce a result no better: centripetal implosion, as the job closes in on itself cerebrally, free of the tangible connection to its outer actions. Thinking is heavy and can wear down the incumbent, while acting is light and cannot keep him or her in place. Only together do they provide the balance that seems so characteristic of effective management.

Too much leading produces a job free of content —aimless, frameless, and actionless — while too much linking produces a job detached from its internal roots — public relations instead of public service. The manager who only communicates or only conceives never gets anything done, while the manager who only “does” ends up doing it all alone. And, of course, we all know what happens to managers who believe their job is merely to control (or, these days, to their organizations, as the detached managers escape in their golden parachutes). If I combine the more cerebral roles, we end up with a “think-link-lead-do” model. Or, as someone who read this description remarked, “Well then, management is a form of thinking and leading by doing.”

A bad pun may thus make for good practice: the manager must practice a well-rounded job. Or, if you prefer, anyone who wants to hold this job must swallow the whole pill (which may not, in fact, be a bad metaphor, since the outer coating of doing provides the immediate effect, while each successive layer in this time-release capsule provides deeper, less accessible, but more long-term effects). The different roles are somewhat substitutable, to be sure, for example, pushing employees through controlling by systems instead of pulling them through leading by culture. But, more important, they are complementary.

In fact, while we may be able to separate the components of this job conceptually, I maintain that they cannot be separated behaviorally. In other words, it may be useful, even necessary, to delineate the parts for purposes of design, selection, training, and support. But this job cannot be practiced as a set of independent parts. As noted, the core is a kind of magnet that holds the rest together, while the communication ring acts as a membrane that allows the flow of information between inner thinking and outer behaviors, which themselves tie people to action.

Indeed, the most interesting aspects of this job may well fall on the edges, between the component parts. For example, Andrew Grove, president of Intel, likes to describe what he does as “nudging,” a perfect blend of controlling, leading, and doing.16 This can mean pushing people, tangibly but not aggressively, as might happen with pure doing, and not coldly, as with pure controlling, but with a sense of leading. There are similar edges between the inside and the outside, thinking and behaving, and communicating and controlling, as we shall see.

Managers who try to “do” outside without “doing” inside inevitably get themselves into trouble. Just consider all those chief executives who “did the deal,” acquired the company or whatever, and then dropped it into the laps of others for execution. Likewise, it makes no more sense to conceive and then fail to lead and do (as has been the tendency in so-called “strategic planning,” where controlling has often been considered sufficient for “implementation”) than it makes sense to do or to lead without thinking through the frame in which to embed these activities. A single managerial job may be carried out by a small team, but only if its members are so tightly knitted together — especially by that ring of communication — that they act as a single entity.

Earlier I mentioned how leading, while identifiable as some activities, tends to infuse almost all managerial work. But the same is true of communicating, as noted, and of conceiving (which cannot be saved for some ethereal mountain retreat), and of scheduling. Likewise, just as the core role of conceiving cannot be separated from the surface role of doing, so too can the inside roles of controlling, leading, and doing not be separated from the outside ones of linking and doing. In fact, at the limit, they blend into one another. Leading and controlling have been described here as inside roles, linking as an outside role. But a manager with considerable influence over an outsider — say, a captive supplier — can exercise leadership as well as direct control, while one whose own employees begin to act as free agents may have to relate to them as a linker more than a leader (as the chief of any clinical service in a hospital can tell you).

Managing in Style

To describe the various components that make up the job of managing as integrated, infused, and well rounded is not to imply that all managers do all of them with equal emphasis. Managerial work does vary, according to the needs of a particular job and the approach of its particular incumbent. Different managers end up emphasizing different things in different ways. Paradoxically, however, this may best be understood through a systematic framework of what they all do in common. Accordingly, I use the model in Figure 9 to consider briefly some questions of managerial style (including the effect of context). Style is considered to impact on managerial work in three ways: which roles a particular manager favors, how he or she performs these roles, and what kind of relationship exists among these roles.

First, and most obviously, managers in different contexts have to emphasize different roles. For example, as noted above, the managers of autonomous professionals, as in hospitals or universities, tend to favor linking over leading (let alone controlling), since professionals tend to come to their work naturally empowered. In other words, they need little encouragement or supervision, although they do require considerable external support. However, when experts must work in teams, as in some research laboratories or in professional sports, leadership becomes rather critical, particularly at the group level. Entrepreneurs, in contrast, who run their own businesses, tend to emphasize doing (alongside conceiving), as they involve themselves deeply in specific issues. Interestingly, the same thing tends to be true of first-line managers, even in conventional big businesses, for example, foremen in factories who must resolve steady streams of operating problems. Senior executives of large diversified firms, on the other hand, give greater attention to controlling, particularly through their systems of performance control and their decisions to authorize major capital expenditures.

Of course, regardless of the context, individual managers are often personally predisposed to favor particular roles or aspects of the job. Considering this from the inside out, we can, for example, distinguish a conceptual style of management, which focuses on the development of the frame, an administrative style, which concerns itself primarily with controlling, an interpersonal style, which favors leading on the inside or linking on the outside, and an action style, which is concerned mainly with tangible doing. And as we move out in this order, the overall style of managing can be described as less opaque, more visible.

Even in how they respond to requests, managers can exhibit subtle yet significant variations in style. Asked for advice by an employee, for example, a manager may respond as a communicator (“Payroll has some data on this”), a controller (“Don’t do it”), or a leader (“How do you feel about it?”). Of course, the doer may say, “Just leave it with me”! All kinds of opportunities arise in managing to substitute, combine, and give nuance to the different roles.

Second, regardless of which roles or aspects of the job a particular manager emphasizes, his or her personal style is manifested particularly in how these are performed. We have already seen this, for example, in different approaches to linking at the extremes (the sieve, the lead, the sponge), and the different approaches to conceiving the frame (passive, driven, opportunistic, and determined styles). Similarly, we saw different predispositions to leading (favoring the individual, group, or unit level). Other obvious variations in style can be delineated according to how each of the remaining components of the model might be carried out.

A final aspect of managerial style has to do with the interrelationships among the various components of managerial work. For example, an important distinction can be made between deductive and inductive approaches to managerial work. The former proceeds from the core out, as the conceived frame is implemented through scheduling that uses information to drive people to get action done. We can call this a cerebral style of managing — highly deliberate — noting that it has been popular from the early POSDCORB writers to the current proponents of strategic planning and bottom-line thinking. But there is an alternate, emergent view of the management process as well, which proceeds inductively, from the outer surface to the inner core. We might label it an insightful style. As Karl Weick puts it, managers act in order to think. They try things to gain experience, retain what works, and then, by interpreting the results, gradually evolve their frames.17

Of course, the preferred approach may well vary with context as well as personal inclination. When a situation is well understood, the cerebral style seems logical, while, under conditions of ambiguity, the insightful style may make more sense. Part of our problem in recent years, in educating managers as well as in their own practice, has been a predisposition toward the cerebral style in situations of increasing ambiguity that require inductive insight. Even “doing” is one step removed from the action: How then is a manager who relies on controlling —three steps removed — to come to serious grips with any real problem in an organization?

Clearly, there is an infinity of possible contexts within which management can be practiced. But just as clearly, perhaps, a model such as the one presented here can help to order them and so come to grips with the difficult requirements of designing managerial jobs, selecting the right people to fill them, and training people accordingly. It is in these areas that we are in greatest need of rich theory, and this kind of description may be one step toward the development of it. In particular, the model may help in the hitherto intractable problem of delineating a useful list of competences that underlie the effective performance of managerial work.

To give some idea of the flesh that might be put on the skeleton of this model, let me describe briefly a few of the managers I have observed. This gives a sense of how context and personal predisposition interact with the various components of the job to define reality.

  • Ann was director of nursing services for a hospital near London; Peter was the manager of a district in the National Health Service of England. The difference in their work was marked. Ann can be described as “managing down,” Peter as “managing up.”

Ann was intimately connected to the delivery of health care on a daily basis, knowledgeable about every imaginable detail of her hospital’s operations. It is people like her who keep the system running. She conceived and she managed outside, to be sure, and her relationships with the autonomous doctors are best described as linking. But mostly she seemed to maintain the internal operations by a certain amount of “doing” combined with a good deal of leading as well as some controlling, all infused with that “spirit of the hive.”

Peter, in contrast, managed largely up, being more concerned with the administrative intricacies of the hierarchy under which he sat than the district over which he ostensibly had authority. District management is necessarily detached from the daily delivery of health care, and so managing inside seemed to be restricted largely to formalized controlling, particularly with regard to systems of resource allocation. I found less “doing” here and not a great deal of leading. Outside linking seemed to be the focus, to maintain good ties with the upper echelons of the hierarchy. As a result, buffering came naturally to this job. If Ann’s style could be described as involved and insightful, then Peter’s was cerebral and deductive.

  • Carol ran a film company in London, producing quality films mostly for television. She could afford to manage neither exclusively up nor exclusively down, although she had to use a different style for each. Trying to capture both the what and the how, I described her work as “hard dealing and soft leading.” Carol seemed to be a doer above all, specifically on the outside, using a vast array of contacts to “do the deals” of film contracts. But once the deal was done, a team had to be assembled, and its work, by competent professionals, had to be not so much controlled as monitored, to ensure that it remained responsive to the client’s needs. Thus Carol’s outside work was focused and intense — “doing” in detail and with a vengeance — while her inside work seemed to be gentler and somewhat more indirect — leading rather than doing. And her overall orientation had to be rather opportunistic, for this is a highly fashionable business. But there was a frame here too, not crystal clear perhaps, but sharp enough to help select the projects and maintain a sense of unity in the company. Carol’s own style would appear to be more insightful than cerebral, and more inductive than deductive.
  • Norman ran a large, Canadian national police force, where he had spent his career. Linking was important: this is a highly sensitive job in which informing and buffering are critical. And so, in his words, he arranged for “no surprises,” especially for the politicians. But above all, he protected and strengthened the culture of the force, which is legendary. This served as his frame, in a way, on which he laid his own liberal stamp. Controlling is obviously not absent from the job of police commissioner, but this one tilted especially toward the collective level of leading.
  • Fabienne was the head nurse of a surgical ward in a Canadian hospital. The contrast between her job and Norman’s is marked, except in its essence — the importance placed on the leading role. There was much less formality in her job and much less movement: she managed on her feet mostly in the nursing station, around which all revolved. If Norman was “on top” of things hierarchically and colloquially, then Fabienne was in the middle of them. This allowed her to solve problems quickly and informally. Linking was relatively less important here; indeed, Fabienne was “not crazy about the whole PR thing,” as she put it. Instead, she could be described as “doing” to lead (in contrast to Norman’s and Carol’s linking to lead), blending her different activities into a central frame that she described as a “caring” style of management: she believed in serving her nurses much as she once served her patients. Management as “blended care” would seem to be most effective, especially in contrast to the all-too-common style of management as interventionist cure!18

These are but a few of the managers in context, but their brief profiles may give a sense of the richness with which the work and the style of managing can be described. We have a long way to go in understanding this most important of jobs. I can only hope that closing this initial circle may be one helpful way to proceed.


1. H. Mintzberg, The Nature of Managerial Work (New York: Harper & Row, 1973); and

H. Mintzberg, “The Manager’s Job: Folklore and Fact,” Harvard Business Review, July–August 1975, pp. 49–61.

2. P.F. Drucker, Management: Tasks, Responsibilities, Practices (New York: Harper & Row, 1973).

3. M.E. Porter, Competitive Strategy (New York: Free Press, 1980).

For a comparison of strategy as perspective and position, see:

H. Mintzberg, “Five P’s for Strategy,” California Management Review, Fall 1987, pp. 11–24.

4. A. Noël, “Strategic Cores and Magnificent Obsessions: Discovering Strategy Formation through Daily Activities of CEOs,” Strategic Management Journal 10 (1989): 33–49.

5. J.P Kotter, The General Manager (New York: Free Press, 1982).

6. Noël (1989).

7. See R. Simons, “Strategic Orientation and Top Management Attention to Control Systems,” Strategic Management Journal 12 (1991): 49–62; and

R. Simons,“The Role of Management Control Systems in Creating Competitive Advantage: New Perspectives,” Accounting, Organizations and Society 15 (1990): 127–143.

8. See H. Simon’s discussion of intelligence, design, and choice, in The New Science of Management Decision (Englewood Cliffs, New Jersey: Prentice-Hall, 1960).

9. H. Gulick and L.F. Urwick, Paper on the Science of Administration (New York, Columbia University, 1937); see also

H. Fayol, Administration industrielle et générale (Paris: Dunod, 1916); English translation, General and Industrial Administration (London: Pelman, 1949).

10. See F.W. Roethlisberger and W.J. Dickson, Management and the Worker (Cambridge, Massachusetts: Harvard University Press, 1939).

11. See Karl Weick’s criticism of my inclusion of leading as a role in my 1973 book in:

K. Weick, “Review Essay of The Nature of Managerial Work,” Administrative Science Quarterly 19 (1974): 111–118.

12. M. Maeterlinck, The Life of the Bee (New York: Dodd, Mead and Company, 1918).

13. L.R. Sayles, Managerial Behavior: Administration in Complex Organizations (New York: McGraw-Hill, 1964);

Mintzberg (1973); and

Kotter (1982).

14. L.R. Sayles, The Working Leader (New York: Free Press, 1993).

15. T. Peters, “The Case for Experimentation: or, you can’t plan your way to unplanning a formerly planned economy” (Palo Alto, California, pamphlet issued by Tom Peters Group, 1990).

16. A. Grove, High Output Management (New York: Random House, 1983).

17. K.E. Weick, The Social Psychology of Organizing (Reading, Massachusetts: Addison-Wesley, 1979).

See also:

H. Mintzberg, “Crafting Strategy,” Harvard Business Review, July–August 1987, pp. 66–75.

18. See H. Mintzberg, “Managing as Blended Care,” Journal of Nursing Administration (forthcoming, 1994).

Reprint #:


More Like This

Add a comment

You must to post a comment.

First time here? Sign up for a free account: Comment on articles and get access to many more articles.