Stop Training Managers to be Bureaucrats

The following is excerpted from Henry Mintzberg's book, Managers Not MBAs.

***

It is fashionable to view the MBA as modern, progressive, change oriented. And it is certainly true that most MBA programs make great efforts to keep current.

In sharp contrast is the image of bureaucracy. It stands for old, staid, change-resistant. MBAs are supposed to be the antithesis of this; indeed, the schools send them out to “bash bureaucracy.”

I believe the reality is exactly the opposite: MBAs who take what they learned about management seriously end up being the bureaucrats.

I am not using the word for shock effect. Bureaucracy has a pejorative meaning and a technical one; I mean both here, quite literally. In the pejorative sense, time and again I have witnessed MBA managers acting like Dilbert’s boss. Not all, but far too many. In the technical sense, classic bureaucracy has two main characteristics: formalization and centralization. MBA programs promote both, and so therefore do many of their graduates.

Control of human behavior through the formalization of activities is the central guiding principle of classic bureaucracy. It is accomplished through plans, systems, and performance measures—all emphasized in MBA education and therefore favored by many MBA managers. To control in bureaucracy typically means to have it down on paper. A market is controlled if a high number appears next to “market share”on a report; quality is controlled if a low one appears next to “defects”; people are controlled if everyone is connected to a boss on a chart; the whole system is controlled if every action is anticipated in a document called a “plan,” from the “strategic plan” on “top” to the most detailed budgets on the bottom.

There is nothing wrong with control, or formalization, or even bureaucracy per se. It is difficult to imagine any organization without some degree of them. Who, for instance, would fly an airline that did not rely on standardized procedures as well as clear divisions of responsibility? The problems arise—and the word bureaucracy takes on its pejorative meaning—when an organization tilts too far in this direction. It becomes impersonal and inflexible, its managers distant and disconnected. In my opinion, MBA education tilts too far in this direction. So do companies run by its graduates who fail to get past their education in their practice, who cannot offset their analytical training with the practices of art and craft.

As for centralization, I know of no MBA program that overtly promotes it. That is neither modern nor progressive. But are there any that do not covertly encourage it—by leaving the impression that managers are important people removed from the context of their managing, people who sit in offices to make carefully calculated decisions and pronounce immaculately conceived strategies for everyone else to implement?

Managers who focus on the numbers instead of the nuances, systems instead of subtleties, especially when parachuted into contexts they do not understand, feel insecure and so grab onto whatever controls are available. The consequence is overcentralization in some areas, through the making of decisions that should have been delegated, and overformalization in others, by expecting systems to control what direct decision making cannot. Such managers are too controlling and too disconnected at the same time, much like parents who cannot get the discipline right and so appear confusingly inconsistent.

Putting this discussion together, I believe that the numbers of MBAs in senior positions coupled with the cult of heroic leadership has made many of our large corporations more hierarchical, more centralized, and more formalized than they should be. All the efforts that developed from the 1960s through the 1980s to involve people more deeply in their work have given way to a callous and arbitrary reign of bureaucratic management. And this in a time, especially in high-technology industries, when we need to favor teamwork, collaboration, and networks.

Consider the popularity of the concept chain in management education and practice. Of course, it is not the vertical chain that is fashionable, that “chain of command” running down a hierarchy, but the horizontal chain, of operations, popularized as the “value chain” in Michael Porter’s 1985 book, Competitive Advantage. It should be realized, however, that this ostensibly new horizontal chain in fact reinforces that old vertical chain.

The idea of depicting business operations as a horizontal chain comes from the most classic sphere of mass manufacturing, the automobile assembly line. Here all tasks sequence themselves in linear order, from the components coming in to the automobiles driven out. Since Porter generalized about the value chain of sequenced functions—from inbound logistics to operations, then outbound logistics, then marketing and sales, finally to services—all sorts of other operations have likewise been seen as chains.

Go into an airport or a hospital, a research laboratory or a project team, and look for such chains. Sure, you will find subprocesses that take this linear form. But the overall activity hardly looks like that. We call airports hubs, shown in Figure 5.1b, for a good reason: they are organized less as sequences of activities than as focal points, to and from which people, things, and information flow. So, too, can hospitals be seen as hubs, not only overall, to which people come, but also at each patient inside, to whom the services flow.

Research laboratories and project teams, however, do not look like hubs, any more than chains, because they tend to have no obvious centers, any more than dominant linear sequences. They are better seen as webs, as shown in Figure 5.1c—loosely interacting networks of collaborative relationships that flow every which way. Depicting them as chains, even hubs, violates their complexity.

Now, where to put management in these different perceptions of organizations? In the chain, that’s easy: on top, removed. As shown in Figure 5.1a, over each link is a manager, and over each manager is another manager. A manager for each and a manager for all! In other words, naturally laid over the horizontal chain of operations is the vertical chain of command. No wonder Porter’s chains are so popular in MBA programs: The manager is on top. Of course, this is how management has been seen for a century, on top. But that has been a century significantly dedicated to mass production. 

In a world of increasing prominence of hubs and webs, however, we shall have to see management differently. Look at the figures and you will see why. Putting a manager on top of the hub or the web is silly: it just removes him or her from what is going on. So where to put the manager?

In the hub, that’s easy: the manager is at the center. In the world of organizations, that change is profound: It puts management in a different place. Seeing oneself at the center, instead of on top, changes the whole perception of one’s managerial world. As Sally Helgesen (1990) put it in her book, The Female Advantage: Women’s Ways of Leadership, women managers “usually referred to themselves as being in the middle of things. Not on top, but in the center; not reaching down, but reaching out” So, for example, to get to strategy in a hub, managers have to reach out, to others; they cannot just deem them down, from on high.

Now, where to put management with regard to the web? When I ask it of people, they hesitate. On top? Obviously not on top—any manager on top of a web would be “out of it.” At the center? There is no center in a web; creating one would “centralize” it and so undermine its free flow of information.

Then the answer becomes obvious: In the web, management has to be everywhere. It has to flow with the activity, which itself cannot be predicted or formalized. But there is an additional answer, less obvious but perhaps more profound: Management also has to be potentially everyone. In a network, authority for making decisions and developing strategic initiatives has to be distributed, so that responsibility can flow to whoever is best able to deal with the issue at hand. If this sounds overstated, then think about the “management” of the Internet.

Put differently, in the web control has to give way to collaboration. “Bosses” and “subordinates” running up and down a hierarchy have to give way to the shifting back and forth between “colleagues” on the inside and “partners” on the outside. Sure, webs need formally designated managers. But to connect and contribute more than to command and control. And that means the managers have to get inside those networks. Not be parachuted in, without knowledge, yet intent on leading the team. No, they must be deeply involved, to earn any leadership they can provide.

***

For more thinking on this topic, check out Henry Mintzberg's book: