Virtual Learning Resource Centre

Top down: Why hierarchies are here to stay and how to manage them more effectively

Book cover

by Harold J Leavitt, Harvard Business School Press, 2005.


This book is not a defence of business hierarchies, but insists that they are here to stay. Our best response is to understand their benefits and disadvantages and to seek to exercise the authority implicit in hierarchies to maximise the advantages. We must balance the humanising and systemising factors to make work life more civilised.

(Reviewed by Edgar Wille in September 2005)

(These book reviews offer a commentary on some aspects of the contribution the authors are making to management thinking. Neither Ashridge nor the reviewers necessarily agree with the authors’ views and the authors of the books are not responsible for any errors that may have crept in.

We aim to give enough information to enable readers to decide whether a book fits their particular concerns and, if so, to buy it. There is no substitute for reading the whole book and our reviews are no replacement for this. They can give only a broad indication of the value of a book and inevitably miss much of its richness and depth of argument. Nevertheless, we aim to open a window on to some of the benefits awaiting readers of management literature.)


Harold Leavitt has a distinguished career behind him and is currently an Emeritus Professor at Stanford. Nevertheless I was prepared to take up the cudgels and resist what I supposed were the implications of his subtitle Why hierarchies are here to stay and how to manage them more efficiently. I have always supported a participative approach to management. Hierarchies don't seem to encourage this approach; so I wanted to see them go and the sooner the better. A number of other authors summarised on the VLRC take a similar view. But in the end I had to capitulate to Leavitt and admit that, whether we like it or not, hierarchies won't go and we had better make them work better.

In the course of his reasoning Leavitt offers a wide-ranging approach to management, while keeping his feet on the ground. He comments at the start that we were telling ourselves that hierarchies were cruel, inhumane, inefficient, authoritarian organisations to be consigned to history. He admits that they have many noxious by-products, such as distrust, fear, conflict, servility, greed and many other debilitating fumes. They are being softened in many cases with more horizontal channels, more teamwork, networking and so on, but these improvements are not supplanting top down, hierarchical architectures. They are here to stay.

The author makes it clear that he is not writing to defend hierarchies; he has encountered too many of their destructive failings. But reality is reality and even though we, as citizens, believe in democracy, yet top down power and control will not be banished from the business world. So he considers why a form of organisation which contradicts our political philosophy is still with us. Secondly he considers whether it matters, particularly to middle managers and what could be done to navigate a way through it.

Back to the top

Alive, well and everywhere

That is Leavitt's description of the present situation, in spite of famous gurus who have written the obituary of hierarchies. The demise of these pyramid shaped, multi-level structures is not even in sight. They may be cloaked in authority-disguising euphemisms, but power is still distributed in unequal ways. People still ultimately have to do as they are told by those "higher up", even if the control is exercised in a more polite way than used to be the case, made necessary by the significance of brains as well as brawn in modern business. But not even the emphasis on intellectual capital can dispense with hierarchies.

The author asks if we are fighting the wrong war when we damn all hierarchies. Should we not rather be seeking to modify and tame them, learning how to live and work meaningfully within them?

The book has three major goals. First it is a reality check to enable us to accept the fact of continuing hierarchies. Second, it is a consideration of how they are changing and whether they have become more compatible with people's needs and aspirations. The third and most important goal is to search for ways of supporting managers who live with the paradox of a softer approach to organisation, side by side with the pressure of increasing speed and the hardening impact of modern technology. How can people still find some personal fulfilment in today's enterprise?

Back to the top

Why we don't like hierarchies

The book goes on to look at the reasons why we would like hierarchies to disappear. Well, they do generate childlike dependency; those who have less power depend on those who have more. The word "infantalisation" is employed. Free to do our own thing outside, but imprisoned within. During our working hours we have to do what others have defined as our duties and responsibilities Then when we leave we are free to make our own choices, if only about how we spend our leisure and when we shall eat.

We see hierachies as shackling us in boring routines, discouraging imagination, quashing creativity, getting us to do useless things and not trusting us anyway. They block warm interpersonal relationships. They may urge collaboration, but they reward competition. All too often they breed greed and immorality, where power corrupts.

Then there are practical arguments against hierarchies. They are inefficient. Activities often continue long after they offer any value, just because they have always been done. We see them as slow, unresponsive and inflexible, though the advantage of size may take them in the opposite direction, and permit risk and innovation. Hierarchies don't really like humans. The more they can get rid of them, the happier are their top management and their shareholders. And even for the top bosses there are big downsides. "Uneasy lies the head that wears the crown".

Back to the top

Why hierarchies persist

In view of all these negative factors why do we still put up with them? It is not enough to blame their continued existence on the bad guys. Hierarchies wouldn't disappear even if the bosses were as pure as snow. After all, hierarchies enable us to feed our families; working for them is better than being unemployed. They also meet our need for achievement; they give us ladders to climb and thus we rapidly get drawn into the system and seek to use it for our own advantage.

Hierarchies also help us to define ourselves; the question at a function "What do you do" normally needs a hierarchy to enable us to give an answer which suits our self confidence. Big hierarchies give us a sense of security, even if this can turn out to be an illusion nowadays. They add structure to our lives. They evaluate us and we like to know whether we are seen as good or bad at whatever we do. While we must conform to the norms, yet if we are given the opportunity to show our individualism, it is by the hierarchy that we shall be granted the opportunity.

Pragmatic as well as psychological forces keep hierarchies in being. In many situations they have an efficiency that would be lacking in a more anarchic atmosphere. Everyone having their place and knowing it, can make for remarkable feats of organisation [such as Trooping the Colour, or getting to the Moon, or tracking down terrorists, EW]. We shouldn't see only the downside. Size can also contribute to the benefits; It is not always inflexible. Organisational life cycles enter into the picture too. A lively egalitarian start up eventually outgrows its stage of "anything goes" and has to introduce some system, which normally means a move toward hierarchy.

Leavitt also points out that hierarchies are the basis of all personal human activity. We prioritise and categorise in our heads and set issues in some hierarchical order, so that we can solve problems.

Back to the top

Humanising hierarchies versus systemising them

The author next looks at how hierarchies have managed to stay alive. A battle has been going on within them for over a hundred years between two streams of thought; the humanisers and the systemisers. The first treats humans as the masters and organisations as instruments in human service. The systemising stream treats organisations as the masters and humans as a resource in the service of organisations.

Humanisers emphasise trust, emotion, empathy, relationships and imagination and look to improve morale, enhance motivation, encourage creativity and increase collaboration, hoping thus to increase the productivity and flexibility of organisations.

Systemisers, on the other hand, focus on logic, consistency, discipline, rigour, facts and the bottom line. To them, unlike the humanisers, diversity is an obstacle to the achievement of good order. They go back to the principles of Frederick Winslow Taylor, his scientific management, his stop watch and precise analysis of work movement. Hierarchy was reinforced by his approach which separated the doing from the thinking and caused the thinking involved in craft work to be diminished. True the workers earned more, but it was a dehumanising process. And so the battle has gone on ever since between the humanisers and the sytemisers.

In the 1950s the humanisers hit back. They accused the systemisers of having adding machines where their hearts should be and the systemisers derided the emphasis on motivation and participation.

Then came electronic information technology with opportunities for new jobs and professions, but also for the elimination of middle managers, which, however, did not come to pass. The increased complexity of business filled the gap. What did happen was increased access to information, which made for direct contact between the top and what was going on down below. Such flattening of hierarchies as took place enhanced the power concentrated at the top.

The battle is seen as misconceived. There are elements of systemising which reflect human love of puzzle solving, exploring and inventing. And a PC on every desk, irrespective of level, can enhance work interest. Moreover humans do like to bring order out of chaos.

Back to the top

Three changes that shook the hierarchy

One side accused hierarchies of infantilising humans. The other asked how you could govern society, produce goods and services and maintain our civilisation without paying "some behavioural bills". The battle did lead to some changes.

Humanisers drove the first of three changes – participative management. Sytemisers exploited new technology to propel the second change – analytic management. The third change was the emergence of what are known as hot groups, who thought analytically, yet tended to thumb their noses at authority. And often in one organisation all three change forces would be at work simultaneously.

The humanising influence lay behind participative management, with people like McGregor of Theory X and Y fame, Maslow and Likert. This moved from the cynical view of the worker as lazy and uninterested to involvement and ultimately empowerment.

But managers began to worry about their managerial prerogatives and trade unions saw some of their power ebbing away. Business schools saw their neat curricula as under threat. The new technologies came and enabled many would be participants to climb the ladder and be part of the higher levels, reinforcing the gap between power and authority and the blue collar workers.

Then came the Japanese approach to management in the quality movement, which involved low level workers in responsibility for action and stressed teamwork, knowledge and respect, yet was very much an analytic and systemising approach.

Thus total quality management and just in time methods, along with information technology, brought in the second change – analytic management, the modern information and control systems. Hordes of MBAs entered the business world reared on analysis and quantification and infected organisations everywhere with the systemising virus, often called "paralysis by analysis" And the two parties were further than ever apart.

Then emerged the third change – the "hot group" phenomena, early known as "skunk groups". They were neither dedicated to numbers nor to "touchy feely" experience. They loved the task and were prepared to devote themselves to it, with scant respect for anything else such as impact on people or the powerful ones in the organisation. They did have an impact on organisations, but eventually they, too, became part of the hierarchy, as revolution slowed down.

The three approaches now often find themselves at work in the same organisation and all "piled into the hierarchy", but not in a coherent way. Yet somehow the overarching hierarchy and the dedicated middle managers "hold together the whole, diverse, squabbling family".

Back to the top

Change and authority

Leavitt continues by looking at the impact of these changes on the people involved and the authority, particularly of middle managers. Middle managers have to implement the changes and get good work done in ever more complicated structures.

"Hierarchical organisations allocate varying amounts of authority to their managers. This authority is intended to serve as a major power tool to help managers influence, control and direct business processes. It is also a tool with which people higher in the hierarchy can hold those below them responsible for outcomes." Give them authority and they can be held accountable for getting the job done. So the thinking goes.

However authority is not something which is just allocated. It just goes with the job; what you are responsible for affixes authority to you, however much you might wish to believe otherwise. This authority, implicit in being a manager at all, sends signals to those below, which affect their expectations and behaviour. No manager can act as if he or she did not have authority. It affects all relationships and even the interpretation put upon casual comments. It might be exercised in a more benign way, but it is still there and always will be as long as things have to get done. It is an inescapable reality of hierarchical life, whether we like it or not.

Many managers may not wish to exact obedience from those who report to them, but power to do so is part of the definition of authority, implicit in hierarchies. Managers are there to get people to do what they want them to do, because they, too, have managers who have to get them to do what the higher manager wants them to do. They may also have the power of rewarding behaviour; there may be carrot as well as stick, but, hidden though it may be, the manager cannot wish the stick away.

Many a new style manager has assumed that he or she operated in a participative mode, only to discover that this put them out of harmony with the expectations of their seniors in the hierarchy, and they had to learn to be more authoritative as a defence mechanism. At our peril, we ignore the fact that hierarchies just are authoritarian systems. That is the atmosphere they compel us to breathe, even if they send their managers on courses about team work, listening and coaching, and obscure the reality.

Performance appraisal systems usually reinforce this authority especially when linked to rewards, where, say, the bottom 10% are to be moved toward the door. And the managers, who fail to evaluate the worth of their subordinates in this way, may find themselves in the bottom 10% in the assessment of their managers. There is a skill, which some managers acquire, to use their authority to limit authoritarianism in their part of the organisation – to minimise the stifling effects of the hierarchy.

Managers do develop an authoritative presence, even if they don't consciously do so. It consists of authoritative savvy, such as getting things done through others, networking and connecting, not always with manipulative style. They have the authority and they know how to use it. Then some develop what the author calls an authoritative aura, which emerges in the ordinary things they say and do, or just from their being around. Some aspects of this are forced upon managers, as when a newly appointed manager has to make clear to an uncooperative member of staff who's in charge. You just can't abolish hierarchy, not if you want to get anything done.

Of course there are informal sources of authority and power, and a wise manager makes use of these, including staying on the right side of the boss's personal assistant, and chatting with his chauffeur if he has one.

Middle managers are balanced on the rungs between the bottoms and the tops of big organisations' hierarchical ladders. They may have great problems in maintaining their integrity and their commitment to personal values in face of the threat of senior displeasure. Where wrongdoing is involved this may be a real time of testing; and to speak out may be dangerous to a career. Of course, the middle level accountant who falsified figures in World-Com Inc at the command of her superior, may have wished, as she stood in the dock, that she had chosen the risk to her career.

Leavitt summarises this chapter: "Hierarchies are not what they used to be. They've been somewhat flattened, networked and teamed. Nor is authority what it used to be. But it's still there and it's still the central reality of organisational life."

Back to the top

Leadership, authority and the hierarchy

The job of a manager in finding a path through all the contradictory factors we have considered, is so difficult that perhaps managers can no longer be simply managers. Maybe they should be leaders, which is the OK word of today. Leaders are supposed to be able to work their way through hierarchies which have gone participative, analytic and "groupy" all at the same time. Managers with authority are converted into leaders with the charisma, persuasiveness and vision to get the difficult task done.

And so courses on leadership abound, in the hope of achieving a metamorphosis of managers into leaders. What are the differences? What is a leader anyway? Amazon lists 70,000 books about leadership! Some say that managers are people who do what their job description prescribes. But leaders are their own people, independent, autonomous, and beholden to no one. Is that really true? Certain it is that managers need other tools besides authority. Does leadership fill the bill?

Defining leaders is difficult as they come in so many shapes, and they may not always be virtuous. (Hitler, Saddam Hussein) Three themes emerge from the literature:

  • leaders transform their corner of the world
  • leaders can persuade and induce a group to share and work for the objectives of the leader
  • leaders communicate competence and confidence to people who become their followers.

And of course we don't just follow leaders; we partly create them, because we want someone to look after us and preserve us from danger. This may create toxic leaders who ultimately believe they actually have the powers we are wishing upon them, or who just enjoy the chance to wield power anyway. And there are psychological similarities between the good leaders who successfully respond to our demands and the bad leaders who deliberately manipulate us when they see the chance we are offering them.

Managerial authority will be implicitly accepted for the majority of the tasks needing to be carried out, but when it is necessary to break out of the usual bounds, that's when leadership comes into play – persuasiveness, inspiration and so on. Perhaps for 90% of the time people don't need leadership. And surely in that 10% where leadership is required, it may work only because the manager, turned leader, has authority from the hierarchy.

Perhaps the difference between managerial authority and leadership is that leadership is not something that organisations can give to their managers and then take away. "The qualities of leadership belong to the person and not to the organisation". Leadership qualities can be developed, but the organisation that is paying for the process will not own the product.

Many writers consider that there are clear distinctions between leaders and managers. Warren Bennis has a list of them with items such as: "The manager administers; the leader innovates"; "The manager controls; the leader relies on trust'; "The manager focuses on systems and structure; the leader focuses on people'. [See also The Mind of the manager; the Soul of the leader.] Managers are presented as earthbound, while leaders reach for the stars. "Manager" is just a job title; leaders are seen as having inherent qualities; they are not appointed. Even systemisers look for leadership because of the wide range of interactions involved in a global environment.

The very idea of leadership is indirectly an acknowledgement that hierarchies are unavoidable. Leaders imply followers and leaders plus followers translates into hierarchies – not the old rigid hierarchies, but hierarchies none the less. Even though, in a sense, followers may have ‘elected' their leaders, chosen to follow them, yet once this has happened the leaders have authority.

They will exercise this authority, and will do it in line with organisational needs. And that means hierarchy, softer maybe, but hierarchy still, maybe with less conflict between humanisers and systemisers, with every manager having some leadership elements in his or her role, to help both the followers and the enterprise as a whole to succeed.

Back to the top

A three part model for manager/leaders

The book presents a model for manager/leaders. The model is simple and consists of three steps:

  • Pathfinding
  • Problem solving
  • Implementing.


The discussion of the model starts with the third step – implementing, which is the action part of managing/leading. Research into what managers actually do shows that they lead a very fragmented work life, reacting to issues others bring to them and actually doing something about them. But the more leadership elements are injected into the job, the more they do things through others, and have to depend on influencing and persuasion. However, ultimately it is their authority that makes sure that the leader's wishes are carried out, even if this is only occasionally overt.

People will do 90% of what the manager asks of them, just because he or she is the manager, but the 10% where subordinates are less readily compliant, requires the humanising qualities of leadership; these usually involve some understanding of the emotional component in relationships. Leadership is still operating to achieve what must be achieved according to the hierarchical framework, but the handling of feelings comes to the fore, for which normal hierarchies are not well equipped.

Hierarchies are designed as rational machines which respond to instructions from above and are emotionally neutral about them. Because of this, eventually the hierarchical approach had to allow into its behaviour repertoire the non rational recognition of human emotionality. Things like listening, empathy, encouraging, and supporting have to be exercised – which from a purely rational point of view may be regarded as a waste of time.

Such social influencing is best learned from experience, though a few business schools have moved on from purely analytic instruction to providing help in the emotional areas. [Goleman's work on emotional intelligence has also received much attention recently.]

Problem solving

The second part of the model is about problem solving, which is more about rational and logical issues. While it still aims to influence people, this step proceeds from a different perspective. Problem solving collects the facts and makes a decision, then implements the decision, by hierarchical authority and formal controls. However even here it is found to be wise to involve the implementers in the decision process, to avoid alienation in the implementation phase.

This is where the real skill of manager/leaders is required. Above them they have the number crunchers and planners, who set unforgiving budgets and absolute deadlines. You have authority, so you ought logically to be able to implement, say the bean counters. So humanising methods are essential, but the implementation still takes place within a predominantly systemising, authoritarian hierarchy. And the humaniser will have to pay attention to words like rationality, rigour and discipline.


Equally important is the first of the three parts of the model – pathfinding. You need to identify the right problem that needs tackling. Pathfinding is about fuzzy, but critical, ideas, such as vision, values, imagination and determination. It is therefore very much about the manager/leaders themselves. It is about the direction in which the leaders intend to lead. Pathfinding is not about forecasting the future, but about designing it; about choosing where you want to go. It goes beyond the rational "get the facts, pilot experiments, figure out the return on investment of conventional problem solving".

Pathfinding is not for individuals alone. It requires, but does not always seek or get, the cooperation of those who are going to implement, so that they are sold on the underlying intention and can see a path through the forest. It links with the other two parts of the model. To be just a visionary solo pathfinder is not enough. Most pathfinders are not solitary individuals and they share the stages of their visions.

Back to the top

Integrating the model

As middle managers are the core of implementing activity, they may find that the top managers think of themselves as the sole pathfinders and they will expect to dictate the path. So manager/leaders need to recognise that they are embedded inside an authoritarian hierarchy, and that their personal visions are unlikely to reshape that in a short period. Nevertheless they should not "park their soul outside the gate when they go to work in a hierarchy".

Moreover they have to learn how to cope with the convergent tendency of problem solving as well as the often divergent tendency of pathfinding. You also need to bear in mind the problem creating element in pathfinding, before problem solving is tackled.

So, hemmed in by the hierarchies constricting regulations, the manager/leader is faced with the dilemma of how to make the three elements of the model work together. It is going to take a lot of negotiating, compromising and intermixing to achieve an integrated approach. Pathfinding and implementation have some common ground, but in between comes problem solving which fits less naturally.

Back to the top

Hierarchies, managers and morality

Managers are generally in a state of imbalance in relation to authoritarian hierarchies. They are subordinate from the perspective of their bosses, but wield authority as far as their subordinates see it. They are on both sides of the power relationships and may find this difficult to live with. They may lose sight of their duties beyond the actual task in hand – "duties to colleagues, to employees, to society at large and to themselves".

Leavitt exhorts managers and manager/leaders never to forget that authority will surely be everywhere around them, and on them as well. They can't wash it off. They may prefer the non authoritarian way, but their managerial authority is a force that will always be with them. And when the chips are down this will be evident, however unwelcome.

We are brought up to believe that bullying and coercion is wrong, that people must be treated as equals, irrespective of gender, ethnic origin, or class; we learn to appreciate diversity and to grow out of dependence. Then we go to work and discover that most big organisations don't play by these rules. As we grow in experience, we may even learn to make benign use of our authority and we may also realise that the authority with which we are invested is most effective when least used.

The author neatly suggests that we have to be aware of our authority, but also to beware of it. Exercising direct authority, pulling rank, is likely to be counter productive, reducing the scope for collaboration, connectivity and persuasion. It is also often difficult to know whether the team you lead are understanding and believing in the path you are helping to mark out or whether they are just being obedient.

Effective leadership can be confirmed by looking beyond the direct authority relationship to the horizontal activities. Effective leader/managers will be persuasive, imaginative and convincing as they cross boundaries and work with other functions, with customers, suppliers, the community and society. They will be connective leaders. They will be competitive when necessary, but they will not set out to beat others; they will not work on the principle that their success requires the failure of others.

Connectivity will not obliterate the hierarchy. The hierarchy will remain "the base camp" from which the leader moves out to connect and collaborate or even to compete, with minimal help from formal authority, but a recognition that the hierarchy's controlling eye will still be watching.

Authority implies accountability and still reaps rewards and punishments for the leader/manager. But he or she also has duties to those being led or managed. Authority means responsibility to them and for them. It is not "a gift of personal power". Leavitt comments on the low rating given to these wider responsibilities in business and in management education. "The graduating MBA, regrettably, takes no equivalent to the MD's Hippocratic oath." Rather MBAs tend to come out of this education with their eyes hypnotised by the bottom line, with "numbers in their heads, but little compassion in their hearts" [Ashridge and a few other business schools have deliberately set their faces against this approach in their MBA programmes. Henry Mintzberg is seeking to increase the number of schools following this perspective.]

The book closes with an emphasis on the duty and responsibility that managers and leaders owe to themselves, to hold on to their values and not to allow authority invested in them or exercised upon them to cause them to deviate from civilised ethical standards.

This book stresses the inevitability of authoritarian hierarchies and recognises that "hierarchies remain the most workable and effective structures humans have yet invented for performing large, enduring, complicated tasks" whether in complex manufacturing or political/community governance.

Yet in his closing pages, Harold Leavitt calls upon us to keep needling the hierarchies to become better citizens, closer to society's espoused values. "We can keep pressing them…… to focus more on humanising, even as they continue to systemise".

He continues: "Senior executives, perhaps even more than their juniors, need to be reminded again and again that their positions entail far more than the mere maximisation of stockholder value." "So whether we are high or low on our hierarchical ladders, let's behave like civilised human beings. Let's make our organisations – hierarchical though they may be – more enriching habitats for other responsible civilised human beings."

Back to the top