from Brian Martin, Uprooting War (London: Freedom Press, 1984); this is the revised 1990 version.
Go to

Index page for Uprooting War

Brian Martin's publications on peace, war and nonviolence

Brian Martin's publications

Brian Martin's website

For many people the word 'bureaucracy' conjures up an image of a mass of office workers buried in mounds of paper and tied to a set of petty rules, the notorious 'red tape.' Bureaucracies are often the focus of popular dislike, especially because they are perceived to be inefficient and lack flexibility to meet individual requirements. The infamous 'they' who are continually meddling in people's lives are often thought of as remote bureaucrats.

While the popular perceptions about bureaucracy reflect some insights, they are not a good basis to begin analysing a social structure. To do this bureaucracy needs to be looked at as a set of relationships between people.

Bureaucracy is a way of organising work in which people are treated as interchangeable and replaceable cogs to fill specialised roles. Two key features of bureaucracy are hierarchy and a specialised division of labour. Other characteristics of an 'ideal' bureaucracy are rules which describe the duties of members, a set of standard operating procedures, and impersonal relations between members. In a model bureaucracy, initiatives and policy directions come only from the top echelons. Work in carrying out policies is done at the lower levels within the guidelines set from above.

Most large modern organisations are bureaucratic in form: government departments, corporations, political parties, churches and trade unions. None of these real organisations are pure bureaucracies. For example, initiatives and policy directions in political parties and trade unions sometimes come from the rank and file.

Bureaucracy and the war system

How is bureaucracy as an organisational form connected with the modern war system? To begin, most modern professional military forces are run as model bureaucracies. This is an important connection. But although most wars are fought by armies, they are fought on behalf of states, and bureaucracy is the key organisational building block of the state.

The state is composed of numerous bureaucracies at national and local levels, to administer policy concerning government finances and taxation, the military, economic production, law, transport, communications, etc. If industries are run or regulated by the state, this operation is usually organised bureaucratically. Most services run or administered by the state, including schooling, medical and welfare services, are handled according to bureaucratic principles. Indeed, with few exceptions it may be said that the modern state is made up of bureaucracies.

Does this mean that bureaucracy is necessarily a linchpin of the war system? To obtain a better view on this it is useful to examine the history of bureaucracy and the state.

There are a number of examples of major bureaucracies in ancient times, such as the pyramid-building 'armies' of slaves in Egypt under the Pharaohs, the mandarin system in ancient Chinese empires, and similar forms of rule in various forms of so-called Asiatic despotism. Some of these systems were extensive and rigid bureaucracies. It should also be noted that ancient bureaucracies were usually associated with dictatorial political systems. That bureaucracies were found so useful in these arenas is suggestive of the future of this organisational form.

The expansion of modern bureaucracies occurred in conjunction with the rise of modern states and of professional military forces in service of the state. The key events occurred in Europe in the past several hundred years. According to Henry Jacoby in The Bureaucratization of the World, the rise of bureaucracy occurred as the ties to local groups weakened. The feudal system was based on considerable local economic and political self-reliance. There were many local centres of power, including the church, estates, local aristocrats and provincial centres. People had close ties and psychologically identified with family, land, manor and church. All these aspects of the feudal system were resistant to the extensive division of labour and centralised control required for the operation of bureaucracy.

The feudal system was based on severe inequality and exploitation, and on a narrow physical and mental world which permitted little scope for oppressed groups such as peasants and slaves to organise for change. The feudal estates were also quite warlike. Because there was no higher lord to which appeal could be made in the case of disputes, bitter and prolonged private wars between fief holders were not unusual. Such wars were possible because the coercive power used to control serfs and peasants within estates could also be turned against external opponents.

The local self-sufficiency and autonomy of the feudal system began to break down under the impact of increased trade and commerce, both in goods and in ideas. Towns became centres of independent enterprise, and also provided niches for independent thinking and challenging of religious dogma. The towns, to obtain independence of the feudal lords, looked to the king, hitherto only a leader among equals, for support.

Once the economic self-sufficiency of the feudal domains was eroded, the stage was set for the rise of state power, often under a monarch. A key to the power of the monarchy was taxation. To impose taxes not only on towns but also on feudal estates, hoards of tax collectors (bureaucrats) were employed. Bookkeeping and administration were also required, and the state bureaucracy grew apace. One of the important avenues for expansion of early state bureaucracies, for example in France and Prussia, was to provide training, supervision and supply for large military forces. The state, once it gained significant power over the feudal landowners, used its economic and military power to further destroy sources of resistance: trade monopolies and regulations were established and central police and prison systems were expanded.

To enforce its powers, the state relied ultimately on military force. With its ever-growing power of taxation, larger armies could be maintained. The army consumed a large fraction of state finances. Armies remained mainly mercenary until the French Revolution, in which popular support and involvement in military forces was mobilised for state goals. By this time the role of bureaucracy as the organisational form for administering state power was well established.

As the feudal system declined, so did feudal warfare, including its 'polite' forms such as duelling. With the decline of feudal warfare came the rise of modern war, organised around the modern state, bureaucracy and military. Feudal and modern warfare each reflect the use of organised violence to protect the interests of dominant social groups.

This thumbnail history omits most of the detail and complications of the development of the connection between bureaucracy, the state and the military. But it does suggest the strong connection between bureaucracy and the modern war system. In particular, bureaucratic organisation allows the central administration of large areas of life necessary to maintain and expand state power and its monopoly over mass violence. In addition, the organisation of society along bureaucratic lines serves to destroy independent sources of economic and political power.

To what extent is bureaucracy as an organisational form a root of modern war, and to what extent is the problem simply the directions to which bureaucracies are turned? In other words, can bureaucracy be reformed or must it be abolished or transformed out of recognition? It is important to sort out thoughts about this issue before launching into campaigns to change bureaucracy. What precisely should be the goal of such campaigns?

From one perspective, the problem is the uses to which bureaucracy is turned. Bureaucracies after all can be used to enforce environmental protection and provide welfare payments to the poor as well as to run wars and spy operations. However, this view of bureaucracy as intrinsically neutral is flawed because it does not address the issue of which groups in society are in a position to 'use' bureaucracy. Bureaucracy thrives much more readily in systems of centralised power, not surprisingly considering that bureaucracy is based on the principle of hierarchy. Direct democratic control of bureaucracies is almost a contradiction in terms. In liberal democratic political systems, the most that can be claimed is that state bureaucracies are controlled at the top by elected representatives of the people. Even this so-called popular control is implemented seldom enough. In practice, state bureaucracies in capitalist societies are strongly influenced by corporate elites via provision of jobs, perks and most basically by providing a reason for the state bureaucracies to exist.

In authoritarian political systems, there is less pretence that state bureaucracies are controlled by the people. It is not for nothing that bureaucracies have been prominent not only in Asiatic despotism in earlier times but also under Nazism and Stalinism in this century.

But as well as being a tool for certain class interests such as capitalists, bureaucracies serve their own interests, especially those of the bureaucratic elites themselves. It is typical for bureaucrats to stick by procedures even when this wastes enormous amounts of resources, to tightly control information, and to not tolerate internal dissent. These are all parts of a general defence of bureaucratic interests.

If one insists on seeing bureaucracy as a tool, then it should be seen as a tool easy to use by elites and very difficult to use by any group practising self-management and direct democracy. Bureaucracies are no more neutral tools than nuclear weapons are neutral forms of technology. Bureaucracy is both designed for and selectively useful for a society based on inequality and centralised control. Being prepared for modern mass warfare is one of the ways in which such a society maintains itself. Bureaucracy is therefore not only implicated in serving the war system, it is a mainstay of the system itself. To remove bureaucracy as a root of war, it would need to be restructured along the lines of self-management. With such a thoroughgoing transformation, the result could scarcely be called bureaucratic.

I turn now to a closer look at bureaucratic organisation itself and then to some grassroots strategies for transforming bureaucracy.

The nature of bureaucracy

As mentioned before, bureaucracies are characterised by hierarchical authority, a detailed division of labour, a set of rules and standard routines, and impersonal relations between staff. Not all bureaucracies will manifest these characteristics to the same degree. Here I will approach bureaucracy as a political system which facilitates elite control.

It is useful to compare bureaucracy with the factory system of production. Stephen Marglin has analysed the origins of the industrial revolution. The earlier system of production was the 'putting-out' system: workers produced goods for the market in their own time and under their own control, commonly doing the work at home. Capitalists might handle raw materials and also retail distribution, but control over the speed and method of production remained in the hands of the workers. The factory system grouped these workers together in supervised workplaces. According to Marglin, this did not initially increase the output of goods for a given input of materials and labour. The same production methods were used. (Labour-saving technological innovations came after the establishment of the factory mode of production.) In fact overheads in equipment and supervision were higher, so overall production efficiency was lower than with the putting-out system. But the factory system allowed capitalists greater control: they could force workers to work longer hours, and were able to control the output more tightly. The capitalists increased their profits and used this to extend their control.

Adam Smith used the example of pin manufacture to argue that the factory division of labour increased efficiency greatly. Marglin has exploded the logic behind this example by showing that the increased efficiencies of the division of tasks (drawing the metal, straightening it, cutting it, pointing it, grinding it, etc.) do not require a corresponding specialisation of labourers. The tasks can just as well be done by the same person, one after the other. The manufacturing division of labour is only one way to organise production. It is a way that reduces the control workers have over their work. Marglin thus has shown that the driving force behind the introduction of the factory system was not increased efficiency at all, but the greater control it offered to capitalists.

Bureaucracy, like the factory system, is a way of organising workers. The factory system organises manual workers. Office bureaucracy organises mental workers. Both the factory and bureaucracy are commonly justified by their alleged efficiency. Some factories and bureaucracies are efficient in certain senses, others are not. But the driving force behind bureaucratisation is not efficiency. The key to both the factory system and bureaucracy is that they are organisational forms which facilitate centralised control by elites. In both cases this control is enabled by hierarchy and a fine division of labour.

Fred Emery argues that the key to bureaucracy is the location of authority and responsibility for coordination at least one level above those who are doing the work. The division of labour is not an evil if it is arranged by the workers themselves. It is the combination of hierarchy and the division of labour that allows control by elites.

Rather than seeing bureaucracy as a form of organisation designed for efficient administration, bureaucracy is better understood as a political or power system. Top bureaucrats have the greatest formal power. The hierarchy and division of labour also permit powerful outside groups to have a great deal of influence, including corporate or other bureaucratic elites in capitalist countries and communist party elites in communist countries.

Deena Weinstein in her book Bureaucratic Opposition has developed most effectively the idea of bureaucracy as a political system. She argues that bureaucracies are analogous to authoritarian states: in both cases people are expected to stay in their places, to do as they are told, to offer opinions only when asked, and to identify solely with the rulers and the official ideology. Within authoritarian states, and within bureaucracies, individual and collective oppositions exist. The opposition may be to particular policies, to corruption, to exploitation or to organisational structures. Rather than being misfits who are disturbing efficient functioning, bureaucratic oppositions should be analysed as political oppositions, that is as challenges to the use or distribution of power in the bureaucracy.

Weinstein's analogy between bureaucracies and states is particularly revealing with regard to their links with the war system. Bureaucracies and states each prop up systems of privilege and power. It is appropriate that bureaucracy, as the building block of the state, is similar in the nature of its power structure to a state, an authoritarian state no less!

One important difference between bureaucracies and states is that most bureaucracies rely only on nonviolent sanctions against dissidents, whereas states can call on police and military forces if necessary. Most bureaucracies rely not on the use of force but more on a system of rewards, including favourable feedback and promotions, and on a system of rules that legitimises the structure. Willing service to 'higher causes' within a bureaucracy or in a state provides much more stability than reliance on coercion. Antagonism is further subdued by permitting nonconformity within limits, and using various methods to buy off discontent and coopt dissident leaders. Non-coercive control is all the more effective because it is difficult to recognise and to oppose.

Under state socialism the dominance of bureaucracy is quite overt. State bureaucracies administer all possible aspects of life. In parallel with these state bureaucracies, penetrating them, controlling them and constrained by them is another powerful bureaucracy, the communist party. In each case bureaucratic elites are in positions of state power. Hence state socialism is also sometimes called 'bureaucratic socialism.'

In capitalist societies the dominance of bureaucracy is less immediately evident, but the practice is not vastly different. In many capitalist societies, national economic and political directions are set through a system which is called corporatism. Elites from key influential sectors, typically government, corporations, state bureaucracies and trade unions, get together formally or informally to negotiate the framework for political and economic decision-making. This may occur through national planning agreements between corporations and trade unions, by creation of government departments or advisory bodies on women's affairs, the environment or science, or bipartisan agreement on military expenditures.

As I interpret it, corporatism is essentially coordination by elites, most of whom are bureaucratic elites. To have an effect on policy, one must work through a bureaucratic structure in one sector or another, whether it is a political party, a corporation, a trade union or an environmental advisory body. The appearance is that all interests are represented. The bureaucratic underpinning of corporatism ensures that power remains at the top.

Bureaucracies both incorporate and mobilise other power structures. Men in bureaucracies can use their power to exclude women and hence maintain or extend patriarchal power. At the same time, bureaucracies mobilise the power of men over women to maintain the bureaucracy itself: men support bureaucratic power since it is a means for maintaining power over women. In a similar way, bureaucracies dynamically interact with other structures of unequal power, including capitalism, racism and the state.

How can bureaucracy be replaced by a different organisational form which is more participatory, less hierarchical, more responsive to community interests, and generally less easy to be directed towards maintaining or promoting inequality, domination and war? A big question! More immediately, what has been done towards learning how to transform bureaucracies in the direction of self-management? Here I will describe three different approaches towards this goal: academic promotion of and facilitation of industrial democracy, the workers' control movement, and experiences of social action groups.

Academics and industrial democracy

The academic community is quite undemocratic, hierarchical and riven by competition, jealousy and power plays. The status and privileges of academics depend heavily on their position as professionals and their links with other professional groups and managers, all of whom help establish the framework for managing employees lower in the pecking order. The status and privileges of academics are based on claiming areas of knowledge as the exclusive preserve of professional experts. Academia is essentially a competition for power and status carried out using bodies of knowledge as bargaining chips.

The nature of academia helps explain why hardly any of the numerous scholarly outpourings on bureaucracy are useful to social activists. There are many studies of how to control and use bureaucracy, but always from the point of view of those at the top. There is very little material on alternatives to bureaucracy and on how to go about changing bureaucracy from the bottom.

The area of academic study most directly relevant to bureaucratic change is industrial democracy. Academics by and large have ignored or been hostile to this area. When interest has been shown, it has mainly involved study and critique at a distance, and not active involvement in learning how industrial democracy might be fostered. When academics study industrial democracy, it is as something 'out there': industrial rather than academic democracy.

In spite of all this, there have been a small number of academics who have bucked the tide and not only studied but also promoted industrial democracy in the course of studying it. (Actually, to call these researchers 'academics' may be a bit unfair, since many of them are closer to being social activists in background experience and orientation.) One of the main such groups was associated with the Tavistock Institute in the 1950s and 1960s and involved people such as Eric Trist, Einar Thorsrud and Fred Emery. These researchers studied existing and spontaneous examples of autonomous work groups, and drew conclusions about the difference between this mode of work organisation and the usual bureaucratic mode. But they realised that to learn more about the dynamics of autonomous work groups, the conditions for their survival and the sources of resistance to them, they needed to help design autonomous work groups. In practice this meant they were promoting industrial democracy. For an academic to create an experimental innovation, whether a weapons design, a surgical technique, a mode of social analysis, or a form of social organisation, is quite often in effect (if not in intention) to promote it as well.

The Tavistock researchers entered a number of work situations seeking to introduce trials with different forms of work organisation. They sought permission of all parties concerned: management, trade unions and workers. Given permission, the investigators studied the entire work situation: not only the hierarchy and division of labour, but also the technical equipment, the skills required, and the objectives sought by workers and management. They investigated alternatives and eventually proposed a reorganisation of work relationships and technical organisation. Because social and technical factors were intermeshed, this approach is called socio-technical design (or redesign).

To take only one of many possible examples, at a pulp mill in Norway a reorganisation of work involved an upgrading of skills and a limited form of job rotation. The results included improvement in quality and costs of production, better communication and teamwork between operators, and many suggestions from the workers for technical improvement.

From a social science perspective, the results of several decades of experience with socio-technical design are remarkably clear-cut. The evidence shows overwhelmingly that reorganisation of work to increase participation, promote sharing or rotation of tasks, and reduce hierarchy results in equal or greater productivity, increases job satisfaction, reduces absenteeism and increases quality of output.

The alternative is at hand, and it works! The industrial democracy researchers had hoped that socio-technical design innovations, which they introduced in small working groups, would be taken up throughout the enterprises and copied elsewhere. By and large this expectation has not been met. The innovations, however successful, have mostly remained isolated changes or have even been reversed. There are several reasons for this, including the conservatism of management, trade unions and staff and the lack of further injections of the special attention which had been lavished on the experimental groups.

One of the key problems inhibiting further expansion of industrial democracy was that the academics themselves were responsible for too much of the redesign process. In some cases the redesign was largely worked out by the academics. But even when they mainly served as facilitators for redesign efforts by workers, there was no incentive or participation created at the middle management level for expanding the scope of the redesign.

The next step towards fostering a self-sustaining process of work redesign in the direction of industrial democracy has been described by Trevor Williams in Learning to Manage our Futures. He concluded that middle management needed to become active promoters of industrial democracy. Williams helped organise a programme in which middle managers of Telecom (the Australian government telecommunications bureaucracy) in Western Australia attended workshops run by him and his colleagues. The managers clarified long-term goals of Telecom in the light of changing global circumstances. They were introduced to concepts of work redesign and were encouraged to develop projects to encourage workers at lower levels in their own sections to undertake work redesign. The idea behind this approach was that the managers themselves would become committed to a process of organisational self-evaluation and change in the direction of industrial democracy. The approach has shown positive results.

Glenn Watkins has carried forward the work with Telecom Western Australia, and is currently taking it a step further down the hierarchy, getting lower-level managers to involve their subordinates in active decision-making. Economic and political pressures are important here: flatter hierarchies, multiply skilled workers and restructured organisations are ways to increase efficiency, as conventionally assessed. To accomplish this, workers need to be involved in the redesign process. But the terms 'industrial democracy' and 'redesign' are not being used to describe what's happening.

In another experiment, Williams attempted to introduce increased self-management in his own commerce courses at the University of Western Australia. He made an important finding. Those students who were most resistant to collectively organising their learning in a cooperative fashion were those with the longest and least interrupted experience in orthodox educational institutions .

As well as the length of learning, also important was what the students had learned about learning. Some had learned to prefer a stable bureaucratic environment, while others preferred a ruthless competitive struggle. Only a few learned to actively control their learning environment to maximise learning for a changing environment. This suggests that traditional schooling as a form of social organisation is quite contrary to the willingness to engage in self-management.

Starting further down the educational ladder is one response to this problem. Glenn Watkins has applied the redesign process to a variety of organisations with considerable success, including primary schools.

There are many things to be gained by active experimentation in industrial democracy such as undertaken by the Tavistock researchers. Unlike spontaneous efforts towards workers' control, researchers can choose their situation carefully and systematically examine the factors favouring and hindering industrial democracy. Academics are not automatically identified with a particular interest group, and thus sometimes can gain the support of management, unions and workers to promote sociotechnical redesign.

There are also some serious limitations to the academic work so far which seeks to determine how industrial democracy can be promoted. One limitation is the basic justification for increased industrial democracy, which is the need for bureaucracies to survive in a 'turbulent' organisational environment in which traditional methods of top-down control are inefficient or counterproductive because external conditions change too rapidly. Although this perspective is quite useful, it does not provide a basis for fostering industrial democracy when the requirement for 'active adaptive organisational learning' is not so pressing as to demand changes in bureaucratic structures.

Another limitation of academic work is that the issue of recalcitrant or uncooperative bureaucracies has not been tackled. What should be done when management, trade unions or workers don't want to change? In particular, the academic work has depended on gaining management support or tolerance for implementing socio-technical redesign. This means that the changes made in work organisation have mostly been at the shopfloor level. This is valuable, but how is overall decision making in bureaucracies to be democratised?

In addition, the problem of harmful or unnecessary bureaucracies has not been confronted. Even complete workers' control is not much of a goal if what is being controlled is a tobacco company or a nuclear power plant. Experiments need to go beyond industrial democracy to worker-community control.

Workers' control

The workers' control movement probably has more experience in directly challenging bureaucratic organisational forms than any other social movement. The more far-reaching aims of the workers' control movement include workers collectively and democratically designing social and technical work arrangements, workers being direct and equal participants in deciding overall goals, methods and policies for productive enterprises, and workers deciding what goods or services the enterprises should be producing. It is clear that these goals are incompatible with the hierarchy and division of labour characteristic of bureaucracy.

The workers' control movement has grown largely out of the immediate experiences and initiatives of workers, and in many cases action has preceded and stimulated theory. It is interesting that there has been relatively little cross fertilisation between the workers' control movement and the academics promoting industrial democracy. One reason for this is that the academics have sought to obtain management support or tolerance for their initiatives, whereas the workers' control movement in many cases has directly challenged the role of management.

Some of the features, strengths and limitations of workers' control initiatives were described in chapter 5 on self-management. Here some of the aspects of workers' control directly relevant to transforming bureaucracy will be spelled out.

Workers' control means many things to many people. It does not always imply doing away with bureaucratic structures. Indeed, the goal of transforming bureaucracy is seldom explicit in workers' control theory and action, and almost never in relation to the war system.

In many cases workers' control is interpreted in ways which coopt rather than mobilise forces for change. In these cases the terms 'worker participation' or 'industrial democracy' tend to be used. 'Worker participation' can be taken to mean representation of workers in managerial decision-making, or various means of consultation with workers. 'Industrial democracy' can be taken to mean greater autonomy or measures for increased job satisfaction at the shop floor level, without any fundamental change in the hierarchical decision-making structures.

There is considerable debate among certain socialists about whether such piecemeal tokens of 'participation' and 'democracy' should be supported or even accepted by workers. On the one hand, such small changes in the direction of workers' control can increase workers' autonomy and may provide leeway for building campaigns for more fundamental changes. On the other hand, concessions from management (or indeed initiatives from management) to marginally improve participation or working conditions often serve to dampen or head off discontent. Limited forms of participation, such as consultation, representation and self-determination of work within limited parameters, may legitimise the prerogatives of management on fundamental matters.

A major problem facing the workers' control movement is the extreme hostility by corporate and state administrators to any major workers' initiatives which challenge the prerogatives of management. Survival in a hostile capitalist-bureaucratic environment is not easy: supplies are cut off, sales channels closed, insurance is not applicable, legal and police powers are antagonistic. The obstacles are even greater for the many workers' control initiatives launched in collapsing industries.

A powerful source of support for workers' control initiatives is the labour movement, and also some social action groups. Another source of support often sought by workers' control promoters is the government, especially when a social democratic party is in office, and especially via nationalisation. Nationalised industry, in which the state is the owner and manager, needs to be distinguished from socialised industry or self-managing industry, in which control is directly vested in workers and community. Nationalisation by itself does nothing for, or is contrary to, the goal of transforming bureaucracy, since bureaucratic modes of organisation are more entrenched within the state than within capitalist enterprises.

In spite of these and other problems and limitations of the workers' control movement, it remains one of the most important movements for challenging bureaucracy: its most radical goals undercut the essential principles of bureaucracy and its grassroots organisational base challenges the elite control which sustains and is sustained by bureaucracy.

Social action groups

For most social action groups, including gay activists, feminists, antiracists, education activists, environmentalists and antiwar activists, bureaucracy as the dominant form of social organisation has not been a focus of attention. Bureaucracy is often not seen at all: it is accepted as part of the social and political landscape. As a result, there have been few campaigns aimed at transforming large-scale bureaucracies.

Action groups that focus on challenging social problems often work through bureaucracy, sometimes eagerly and sometimes grudgingly. Their aim is to change bureaucratic policies, not bureaucratic structures. Groups fighting for the rights of women, gays and oppressed minorities aim to overturn discriminatory policies and to obtain fair hiring and promotion practices and representation within bureaucracies. Environmentalists seek to stop particular freeways or chemical factories, not to reconstitute the basic nature of social decision-making. Experienced activists pass on their knowledge of how to use the state bureaucracies: who are the sympathetic bureaucrats, how to lobby effectively, how to apply mass pressure to influence policy at key moments.

All of this can be quite useful and often effective, and should not be rejected. But working through bureaucracy on the inside, or demanding policy changes from the outside, does little to transform bureaucracy itself. In fact, working through bureaucracy can reinforce the legitimacy and sway of bureaucracy itself. In addition, campaigns oriented towards working through bureaucracy or applying pressure for change at the top tend to become bureaucratised themselves.

Another important orientation adopted by many social activists is towards building self-managing organisational forms for their own activities, such as cooperative enterprises or egalitarian action groups. Self-managing organisational forms are an alternative to bureaucracy. Direct experience in self-managing groups strengthens the sense of community and commitment to social action and also provides understanding and individual strength to resist pressures for bureaucratisation in the wider society. In as much as social movements organise themselves as decentralised self-managing groups, linked by federations and networks, and self-consciously set out to develop and extend such structures, they provide a strong challenge to the domination of bureaucratic forms of social organisation.

Setting out to 'live the alternative' of self-management is vitally important, but it is not enough. So long as self-managing social action groups remain small and isolated, they provide little threat to dominant structures. The military can tolerate, or squash if necessary, a few conscientious objectors or nonviolent groups on the fringes of society. Likewise, so long as self-managing social action groups remain separate from the day-to-day experience of most people working in large-scale bureaucracies, there is little chance that these bureaucracies will suddenly collapse or transform themselves.

In short, most social activists have either worked through bureaucracies or organised alternatives isolated from the dominant bureaucracies. They have not mounted campaigns focussing on bureaucracy as a key social form. I believe doing this should be a top priority for those seeking to remove the roots of war. The field is wide open and there is much to learn.

Some efforts by social activists have resulted in challenges to bureaucratic control. For example, in some education systems there have been struggles for more participatory decision-making and greater community control. Because schooling is only partly bureaucratised, there is more political potential for teacher and community activists to push for local self-management. There is a great need for such struggles to be studied and for political insights from them to be drawn out. Here I will only describe one small example of a social action campaign focussing on bureaucracy.

At the beginning of 1982 Friends of the Earth (Canberra) decided to organise a campaign around bureaucracy. Since its formation in the early 1970s, FOE-Canberra had mainly campaigned against uranium mining and nuclear power, and to a lesser extent at different times on other issues such as forestry and packaging, whaling, and jobs and energy. Attendance at weekly meetings ranged from 2 to 12, averaging perhaps 6. Members attempted to decide on goals and methods in a participatory way, and gradually procedures for attaining consensus became better understood and used. A continual attempt was made to design campaigns and activities to allow participation by all, to share both boring tasks and exciting opportunities, and to provide emotional support and pay attention to group dynamics while pursuing tasks. Thus it is fair to say that FOE-Canberra aimed at organising itself in a self-managing way.

At the end of 1981 we spent several meetings deciding on priorities for 1982. Uranium mining and nuclear power came out at the head of the list, as usual, but bureaucracy also rated highly. Several of us felt that bureaucracy was in some way at the root of many environmental problems. Environmentalists could write letters, organise protests and use nonviolent occupations, for example to oppose the logging of rainforests. But these campaigns, however successful in their immediate objectives, did nothing to transform the forest industries and government forestry bureaucracies, which kept on with their environmentally destructive policies.

At the beginning, even those of us who were enthusiastic about a bureaucracy campaign didn't really know what this meant in practice. It took us nearly 6 months just to work out what we were trying to do. We had discussions and brainstorming sessions, circulated articles about bureaucracy and talked to people outside our group. Several of our members were sceptical. What were we trying to do? Why worry about bureaucracy? What could we possibly do anyway? One thing was clear: in Canberra, the national capital, there were plenty of state bureaucracies on which to try out any ideas we came up with.

One early idea was to interfere somehow with a particular bureaucracy in Canberra, so as to learn how it operated. The idea was to launch a little probe into the organisation and see what happened. Bureaucracies might seem very stable, but no one was sticking pins in them to see if they had sensitive points. One proposed probe was to write numerous letters that would require replies drafted on behalf of the government minister in charge of a department, and so clog up the system. By learning first about the internal dynamics of the department, and then soliciting support from many other social action groups, we could aim at such a goal of jamming up the bureaucracy.

But on reflection this approach seemed to have at least two flaws. First, it would antagonise those bureaucrats who were burdened with the ministerial correspondence. Second, there was no reason to expect that clogging up the system this way would in any way help to transform bureaucracy, or even provide lessons on how to transform bureaucracy.

With the FOE project in mind, I began a search for ideas about transforming bureaucracy. There turned out to be very little to guide us. The academic work on industrial democracy was useful and stimulating, but not directly relevant since we were not academic researchers. The literature on workers' control was valuable, but we were not the workers. In fact, there was nothing at all that I could find about how social action groups, inside or outside a bureaucracy, should go about learning and campaigning to transform bureaucratic structures. Indeed, most of the theoretical perspectives on bureaucracy were pretty useless for a social action group with this goal. The most valuable perspective was that presented by Deena Weinstein of bureaucracy as a political system.

Eventually I came upon a useful idea. In one of André Gorz's articles, on workers' control, he describes the following: In the early 1960s, a British sociologist named Goldthorpe made a detailed study of Vauxhall workers at Luton. Interviewing them separately, he enquired about their feelings concerning work, wages, and their life situation, and concluded that the workers were integrated into the system. A few militant workers obtained a summary of Goldthorpe's report and circulated copies to workers. Shortly after, a newspaper reported on Vauxhall's large profits which were being sent to General Motors in the United States. After this news was also made known to the workers, rioting broke out at the Luton Vauxhall factories, lasting two days.

So although Goldthorpe found that the individual workers seemed to be satisfied, underneath there was a great dissatisfaction and potential for collective action. Goldthorpe's study contributed to the workers' outburst by focussing attention on issues of job satisfaction.

Gorz's account suggested that we might undertake a survey. We could hardly expect or desire to induce a riot in an Australian government bureaucracy, but in other ways a survey seemed a useful tool. It would help us learn more about bureaucracy, would involve us in direct interaction with bureaucrats, and encourage bureaucrats to think more critically about their own situation.

Even after deciding on a survey, it took several months for us to decide on an interview technique, choose the key areas which we hoped to probe and develop suitable questions. We chose open-ended but directed discussion as described by Ferdynand Zweig. We then practised our interview approach on each other and on sympathetic bureaucrats. We also had to decide on procedures for maintaining confidentiality and pick a suitable section of the government bureaucracy. We chose a division of the Department of National Development and Energy, which we thought would neither be overly sympathetic (such as Environment or Industrial Relations) nor excessively hostile (such as Defence or Treasury).

Our results were illuminating to us, though not very surprising. We obtained responses about job satisfaction and the nature of bureaucratic decision-making which tally with the standard knowledge about bureaucracy. We found as expected that very few respondents knew anything about alternatives to bureaucracy, and fewer still had any ideas about how to go about changing bureaucracy to be a more satisfying place to work in and to be more responsive to community interests.

Perhaps more revealing was the reluctance of many bureaucrats to be interviewed at all. After the initial stages we were told by the top bureaucrats that we could not enter the premises for our survey. But we obtained written permission from the head of the department concerned to interview staff about their personal views, so long as it was done outside the building and outside working hours. Even with this written permission, a large fraction of bureaucrats were clearly afraid of being associated with us at all. FOE in Australia has the reputation of being a radical organisation, and apparently it would be potentially harmful to their careers for bureaucrats to even be known to have talked with us. This response made the similarity of bureaucracy to an authoritarian state quite clear.

Eventually we completed an article describing the nature of bureaucracy, insights from our interviews, and alternatives to bureaucracy. At the end of 1983 we distributed copies of this article to members of the Department of National Development and Energy where we had made our interviews.

Our interview project was at most the first step in a bureaucracy campaign. As it turned out, other activities, considered more urgent, took precedence, and the interview project was not followed up. In any case, many such projects are needed. Some may spark creative initiatives or fall on fertile bureaucratic soil, and provide the example and inspiration for further efforts to change bureaucracy.

Strategy to transform bureaucracy

It would be nice to be able to present a coherent and persuasive strategy for confronting and transforming bureaucracy into self-managing alternatives of autonomous working groups, self-reliant communities, federations and networks, drawing on experiences and insights from a variety of successful and unsuccessful grassroots campaigns to change bureaucracy in this way. Unfortunately the information and experience to draw up such a strategy is not yet available, at least not in organised form. No more than a few isolated social action groups have developed campaigns focussing on transforming large-scale bureaucracy as an organisational form. The more important next step in developing a strategy to change bureaucracy is for more groups to put bureaucracy 'on the agenda.'

What I will do here is outline some principles which I think are important in developing campaigns for transforming bureaucracy.

Link insiders and outsiders

Campaigns concerning bureaucracy are much more likely to be effective if they involve coordinated efforts by people both inside and outside the bureaucracy. Insiders know what is going on first-hand: work conditions, power structures, attitudes, avenues for intervention. They can provide valuable information to outsiders, can advise on what tactics might be misdirected or counterproductive, and can sound out ideas informally. Outsiders have much greater freedom to act without putting their careers in jeopardy. They can take overt stands not safe for insiders to take. Outsiders also can have a wider picture of the role of particular bureaucracies, and are closer in tune with community perceptions.

Insider-outsider links help ensure that campaigns are broad based, and prevent polarisation of attitudes. In many social movements, there is a strong tendency to label all those who are involved with oppressive structures as automatically supporters of the 'enemy' and therefore beyond salvation. This can include government bureaucrats, soldiers, police, corporation managers and political party workers. The result of accepting this attitude and adopting polarising methods is that the insiders close ranks against the attack by the outsiders. Any hope of changing the structure, whether government bureaucracy, army, police forces, corporations or political party structure, is squandered. Treating insiders as potential and indeed essential supporters, and building links with them, helps overcome this counterproductive polarisation.

Similar comments apply to insiders. Many workers in government bureaucracies, police forces, political parties and so forth are sympathetic to the goals of outside social action groups, but may see these groups as amateurish and meddling. The tendency is then to avoid contact with them. This allows the outsiders to become more out of touch and frustrated and adopt stronger tactics, thus polarising the situation. It is far more fruitful to build links with the outsiders and help them become more effective. This does not mean channelling the outside actions into bureaucratic avenues, but rather enabling outsiders to be more effective in their own terms, providing a persuasive challenge to bureaucracies while not antagonising bureaucrats needlessly.

The last word here, 'needlessly,' is important. Polarisation is often inevitable in social struggles. The point is to avoid a polarisation which turns too many people into supporters of the oppressive structure.

Building links between insiders and outsiders does not necessarily require close collaboration in 'mixed' groups. Linking between groups and individuals is compatible with 'separatism' so long as no group imagines its own efforts are the only ones required.

People who are both insiders and outsiders at the same time, such as feminist bureaucrats who maintain contact with outside feminist groups, can play a crucial role. They can be a thorn in the side of the bureaucracy by raising challenges internally, and also provide insights to outside groups to make their campaigns more effective.

James Robertson recognises several roles played by different people in social transformation. Some people spend their time developing and carrying out alternative ways of living and working. Others commit themselves to confronting and eroding existing power structures; they may not have the time or energy to develop non-standard ways of living. A third role is that which Robertson calls 'decolonisers': people who work in a bureaucracy and identify with it, but who are prepared to take part in decolonisation, namely helping people over whom they formerly had power to become independent. Robertson argues that people in these and other roles should try to communicate with and build links between themselves, but that they should also expect inevitable conflicts between the different roles.

People do not just fall into a particular role, such as 'decoloniser,' by chance. Social class, sex, personal history and organisational location can each contribute to this. Since Robertson's decolonisers are so important in helping to challenge and change bureaucracies from the inside, the conditions which produce these people seems a crucial area for investigation.

An example of the unfortunate consequences of lack of contact between insiders and outsiders is the familiar antagonism between social activists and police. Calling police 'pigs' and even fighting with them are only among the more extreme manifestations of activist hostility to police, which is mutually reinforcing. While many members of police forces are corrupt, brutal and conservative, not all are. Many are politically aware and often sympathetic to the causes espoused by the protesters who they must guard or arrest. Police are the agents of social control for dominant groups, not the embodiment of evil.

If there is to be any hope of eliminating oppressive social structures, this will involve transforming police forces as structures. There is a need for self-managing methods for neighbourhood security, for alternatives to conventional prisons, for campaigns to undercut the roots of crime as well as redefinitions of crime, and for conversion plans for police forces. To achieve much of this, building links with sympathetic members of the police is an important task. Already this has been done with many civil disobedience actions. Arrangements with police are made so that arrests do not involve violence. This way, more attention can be focussed on the policy or structure being challenged.

Use political methods

There is little prospect of transforming bureaucracy by exclusively using its own methods, in other words by working 'through the system.' Use of standard channels needs to be linked with methods that challenge the bureaucratic way of doing things, and which incorporate the alternatives being argued for.

This principle has the greatest relevance to those inside bureaucracy. One approach to social change is the 'long march through institutions.' This means climbing the existing hierarchical ladders to obtain formal positions of power, where supposedly one can then have some impact on social directions. The trouble with this approach is that the institutions change most of the individuals long before the individuals rise to positions to change the institutions.

If people in bureaucracies want to change its structure, they can begin at once by raising issues with colleagues, studying and preparing critiques, speaking out on relevant issues, providing support for insider dissidents, and being involved in action groups inside and outside the organisation. Although bureaucrats are often afraid of the consequences of being socially active, there is usually quite a lot that can be said and done without jeopardising one's position. In many cases, establishing a history of principled stands and outspoken behaviour allows a person more scope for further such activity. Others learn to expect dissent.

Climbing to or obtaining high positions in hierarchies is not necessarily undesirable for social activists, so long as this is done without sacrificing one's principles. At higher levels, the dangers of compromise and cooption are much greater. But sometimes the opportunities are greater too. Antiwar generals, corporation executives fostering workers' control and top politicians promoting local self-reliance play a useful role in efforts for social change, especially to the extent that they work with social movements and refuse to play all the 'rules of the game' at the top. Activists promoting self-management from high-level positions are in an inherently unstable position: to the extent that their efforts are successful, their own formal power will be undermined. Indeed, a useful criterion for efforts against bureaucracy is whether top-level power is cemented or eroded.

Action groups outside bureaucracies also need to be wary of working 'through the system.' There is a great temptation to use the normal channels, and use them well: writing letters to politicians and bureaucratic elites, lobbying, being involved in official inquiries, presenting appropriate technical arguments, knowing the right people to contact to have things done. Using bureaucratic mechanisms is often valuable, but it holds little prospect of transforming bureaucracy.

In a liberal democratic state where power relations are massively unequal, lobbying is most useful to powerful groups interacting with each other, such as corporations lobbying state bureaucracies. This is because bureaucracies operate on the basis of centralised power, not logic. For groups without top-down control over physical and human resources, lobbying is largely fruitless and hence interactions with bureaucrats are extremely frustrating, rather like going into shops without any money.

In the FOE survey of bureaucrats in the Department of National Development and Energy, several interviewees told us we were approaching bureaucracy in the wrong way. They said we should be couching our arguments against uranium mining, for example, in narrow technical rather than moral terms. They said we should make an attempt to appear much more 'respectable' and competent, and that we should have tried to introduce our survey through official channels. Though this was not stated, the image spelled out for us was of a slick public relations and lobbying group, rather like that of the corporate lobbying groups that routinely interact with the government bureaucracies. This advice came both from bureaucrats sympathetic to our environmental goals and from those unsympathetic. Indeed, almost all respondents implicitly assumed that our aim was the same as theirs, namely to work through established channels more effectively. They could not easily grasp that we were questioning the nature of bureaucracy.

The great advantage of non-bureaucratic, 'political' methods is that they throw people, and bureaucrats in particular, out of their usual routines and generate awareness of the political nature of social issues and decision-making. Letters and articles in newspapers, distribution of leaflets, public statements, demonstrations and occupations are hard for bureaucrats to cope with. Often they are at a loss in the face of such tactics when used by or on behalf of internal dissidents. When students occupy an administration building, or squatters occupy dwellings, the familiar bureaucratic responses are useless and bureaucratic elites may panic or be paralysed. The familiar dynamic of nonviolent action, in which repression generates greater opposition, can come into play.

Link with fundamental goals

For a small number of individuals and groups, an analysis of the problems of bureaucracy will be enough to justify and motivate efforts to transform bureaucracy. But in general it is important to link such efforts to other interests and principles, often ones which are of immediate or fundamental concern.

While raising issues such as free speech or job satisfaction, it is important not to lose sight of the goal of transforming bureaucracy. The objective should be to link together the immediate concern and the issue of the nature of bureaucracy. For example, suppose a government bureaucrat speaks out about cost overruns in military contracting, as US Department of Defense employee A. Ernest Fitzgerald did in 1970 over the C5A transport aircraft. A defence against bureaucratic reprisals could emphasise issues of free speech, rights of due process and the importance of exposing misuse of public monies. These points could be linked to criticisms of secrecy and the role of vested interests in bureaucratic decision-making, criticisms of the bureaucratic power structures which victimise people who expose such corruption, and proposals for nonbureaucratic alternatives to the current structures.

In linking fundamental goals with bureaucracy transformation campaigns, a careful choice of 'fundamental goals' is vital. For example, one possible argument is that self-managing work groups are more efficient than hierarchically organised groups. But efficiency is mainly something desired by elites controlling or benefiting from bureaucracies. Furthermore, 'efficiency' is usually interpreted only under the presupposition that the process is controlled from the top: that is, efficiency within the existing power relations or efficiency in preserving them. In any case efficiency, however measured, should be a secondary consideration to goals such as democracy or overcoming poverty and alienation.

Rather than 'efficiency,' it is better to use principles such as free speech, job satisfaction and production for social use. These principles are each antagonistic in fundamental ways to the political power structure of bureaucracy.

Promote alternatives

Essential to any strategy to change bureaucracy is an alternative structure. What is the alternative? As described in chapter 5, there are many experiences and ideas for self-managing organisational forms, including self-managing work groups, cooperatives, federations and the lot system, all as part of a society with much greater local autonomy and self-reliance. But in spite of the wealth of experience in nonbureaucratic structures, much more investigation and action is needed to develop stable, effective and attractive alternatives.

It is highly productive to formulate critiques of existing bureaucracies in conjunction with spelling out alternatives. Attention to self-managing systems increases awareness of the key systems of control in bureaucracies, while analysing bureaucracy stimulates awareness of the features of bureaucracy that the alternatives need to challenge and transcend. For example, a critique of bureaucracy might focus on the key role of interchangeability of members of the bureaucracy in allowing hierarchical control. This suggests the importance of allowing and encouraging people to develop and use a variety of skills in a self-managing alternative. Conversely, preference for a strong sense of community and personal support in a self-managing organisation can raise awareness of the way bureaucracy isolates people and fragments social relations through specialisation, hierarchy and working on problems formulated by others.

Formulating alternatives is essential in any bureaucracy campaign. If no alternative is offered, dissatisfaction will remain at the level of gripes or be siphoned off through cosmetic reforms. Alternatives help people see bureaucracy as a social product rather than as part of the inherent nature of society. But more than this, alternatives provide a concrete basis for challenges to bureaucracy. An alternative plan, for example including self-managing work groups or a limited introduction of the lot system, can be a rallying point for both outside critics and internal opponents. The aim here is to turn the alternative into a campaign. In this way the goal of moving from bureaucracy towards self-managing structures is much less likely to become sidetracked.

One way to turn the alternative into a campaign is to actually begin behaving according to the new model. A group of workers could decide to share their tasks and decide priorities cooperatively. The whole panoply of nonviolent action can be called upon, and nonviolent action training used to prepare for opposition as well as to practise the alternative. 'Living the alternative' is something that happens spontaneously and more or less openly throughout almost all bureaucracies, especially at the margins: workers sort out their own work-sharing arrangements, formal meeting procedure remains nominal while de facto consensus procedures are used, individual nonconformists are allowed to go their own way. The introduction of technology for social control and of more refined work arrangements are part of a continuing process in which even these margins of freedom from bureaucratic control are controlled or eliminated. In order for the niches of self-management in bureaucracies to survive and grow, they need to be cultivated, understood and consciously promoted.

One key part of promoting alternatives to bureaucracy is spreading skills and knowledge. Bureaucratic elites obtain a great deal of power by controlling information and breaking up activities into narrow tasks. Any action which makes it possible for insiders or outsiders to understand what goes on inside particular bureaucracies, and to actually carry out the full range of tasks, is subversive of bureaucratic control. Spreading skills and knowledge might take the form of sharing job skills with workmates, describing patterns of decision-making to outsiders, writing exposes of bureaucratic functioning, and preparing manuals and training sessions for others who wish to be able to run or dismantle the bureaucratic machinery.

One of the seemingly hardest problems facing a bureaucracy campaign arises when transformation means abolition. In a transition to a self-managing nonviolent world, many present bureaucracies have little that is retrievable. This includes most aspects of armies, advertising agencies and automobile assembly lines. In these cases internal reform is at best an interim measure.

An alternative plan needs to provide both an alternative organisation and an alternative goal. Instead of hierarchy, division of labour and advertisements, the alternative might be self-managing groups to foster intergroup communications. Instead of hierarchy, division of labour and combat training, the alternative might be self-managing groups to fight fires or build roads.

Or should the alternative function be less related to the old one: instead of drafting memos, a combination of growing vegetables, building bicycles and being with children? The answer is not obvious. It may be more 'logical' to propose an alternative goal that uses at least some existing skills. But if the alternative is too similar to the original bureaucratic mission, a reversion to the bureaucratic model may be too easy.

The solution to this dilemma lies in the hands of all those who take steps to transform bureaucracy. The transformation of the organisational form of bureaucracy and of the goals of particular bureaucracies can be carried out hand in hand. As members of bureaucracies and outsiders gain greater collective control and participation, they will be able to question the goals of the organisation. Faith in the social responsibility of self-managing groups of people must underlie any programme for changing bureaucracy from the grassroots. After all, the premise of bureaucracy is that such faith is unwarranted.