Can Business Beat Bureaucracy?

During the past twenty years, in a profound shift, the leaders of large businesses have turned against the models of bureaucracy that long served as corporate ideals. In rhetoric and practice, managers previously focused on establishing clear lines of authority and accountability. Now they often disparage older hierarchical forms of organization; James Houghton, chairman of Corning Glass, goes so far as to say that "the age of the hierarchy is over." Instead, the new language of management increasingly celebrates involvement, creativity, individual autonomy, participation, even "empowering" employees to use their own initiative.

For those who value participation and human development -- indeed, for all those who care about the extension of democratic ideals in the world of work -- this change in corporate thinking is a startling evolution. For two centuries the dream of worker involvement in management has been persistent but almost entirely futile. From Robert Owen's factories to the communes of the 1960s, reformers have sought to create workplaces with greater involvement, equality, and community. But few have had even temporary success, none in the mainstream economy.

Instead, steadily and inexorably, large-scale bureaucracy expanded into every sphere of human life, driving out or dominating familial or communal forms of organization. And the model of organization that prevailed in government and business alike seemed epitomized by the stopwatch-wielding Scientific Managers of Charlie Chaplin's Modern Times.

Social theorists and popular writers have lamented the trend, decrying the soul-deadening routine of bureaucratic work and raising alarms about the world being lost. Until recently, however, the critics of bureaucracy were rarely found in executive positions. The novelty now is that leaders of large organizations see themselves in the vanguard, attempting to create "post-bureaucratic" organizations based on teamwork and collegiality. And rather than viewing these alternatives as jeopardizing productivity, they see them as a way to make their corporations perform better.

This change has given rise to both new hopes and new discomforts. Those who have traditionally opposed big business in the name of participation suddenly find their old opponents insisting that now they are allies. Years ago, before accommodating themselves to the new industrial regime, many unions fought against the efforts to standardize and "deskill" the work performed by their members. Now the same unions are bewildered by managerial efforts to broaden skills and to demand more active involvement by workers.

The Debate

Are these changes real, or is the current ferment in business merely a new style of rhetoric? Is employee participation in the workplace actually increasing? If a new, "post-bureaucratic" form of work is genuinely emerging, is it robust and effective enough to become the successor to bureaucracy? These are the great questions that, in one way or another, much recent writing about the American corporation has been trying to answer.

In the past few years, optimism about change and enthusiasm for it have predominated among leading business writers, such as Tom Peters, whose new book Beyond Hierarchy is due in February, and who earlier wrote, with Robert Waterman, the best-selling In Search of Excellence; Rosabeth Moss Kanter, author of When Giants Learn to Dance and currently editor of the Harvard Business Review; and Alvin Toffler, the futurist widely known for Future Shock. Offering an attractive synthesis of tough- and tender-minded-ness, these observers and many others have argued that in a modern economy increased participation and increased productivity go hand in hand. "Managers of the world," they say, "free your workers from the shackles of hierarchy, and your business shall prosper."

The skeptics are, for the moment, more muted, but their views are nonetheless important. Some see severe problems of efficiency in organizations based on teams of employees working together ("team systems," in the current jargon). Elliott Jaques claims that the problems of American industry today arise, not so much from bureaucracy itself, as from irrationalities allowed to creep into bureaucratic organizations. Alfred Chandler, our leading historian of American business, adds plausibility to the skepticism -- though his own views are complex and qualified -- by demonstrating the past success of big, bureaucratic corporations that triumphed over smaller firms much like the "team systems" being recommended today. And many other writers, such as Henry Mintzberg, see team-based organizations as narrow solutions to particular business problems rather than as a full-blown alternative to the bureaucratic model.

Others see participatory companies as potentially effective, but worrisome. Many union activists, convinced that militant labor solidarity is the only reliable basis for genuine participation, are concerned about the blurring of lines associated with greater employee participation in management. Among academics, a Marxist strand similarly holds to the view that equality of rights can be attained only through class struggle. For both these groups, participatory reforms are a dangerous diversion, a smoke screen for increased management control.

These differences in perspective have not, however, been debated with any clarity. Indeed, the issues have scarcely been debated at all, since writers with opposing views typically appeal to different audiences and rarely address each other. In addition, many basic terms, such as participation, bureaucracy, and teamwork, are used in widely differing ways. It will help, therefore, if we define a few key concepts and distinctions before proceeding.

Participation has historically occurred at two distinct levels: the boardroom and the immediate workplace. While boardroom participation involves policy-making, participation in the workplace focuses more on improving the organizing of tasks or "quality of work life." For the most part, these have been pursued separately, with employers often favoring participation in the workplace but almost never in the boardroom. But now the most enthusiastic proponents of "de-bureaucratiza-tion" tie these two levels more closely together, envisioning commitment to shared policies at all levels of the organization.

A second issue concerns the concept of bureaucracy. In popular use, bureaucracy is often synonymous with red tape" and inherently a bad thing. Social scientists have tried to use the term in a more neutral sense to refer to a type of organization where, for example, offices are separate from the persons occupying them; authority is impersonal and hierarchical; property belongs to the organization; and so on. "Post-bureaucratic" does not refer to a change in all these characteristics. Most writers primarily have in mind a shift from strict lines of top-down control of individual employees toward a looser form of organization, where employees share more decisions, work more in groups, and are more often held responsible for the results of their work than for adhering to a list of bureaucratic rules. But, as I shall explain, contrasting visions of life after bureaucracy lead to sharply different kinds of organizations.

The assessments that I make here reflect not only recent writing on the American corporation, but my own recent interviews with about 190 middle managers in fourteen large organizations, mostly divisions of large manufacturing firms, including AT&T, General Motors, Dupont, Dow Chemical, Pitney-Bowes, Honeywell, Clark-Reliance, and one computer company. I have followed their reactions to structural changes in their companies, as I try to pin down what works and what doesn't in contemporary business reform.

Current Trends in Management

A few trends have now gone far enough that their basic outlines can be drawn with some confidence. The most familiar, and controversial, developments concern the many forms of "involvement" of shopfloor workers. Equally important are transformations in the work of middle managers. Though the two levels are often treated separately, they are closely linked.

Participatory reforms in blue-collar jobs have been gaining momentum for over thirty years. The most widespread have gone under the banner of "Quality of Work Life" ("QWL" for short). These efforts bring together groups of workers to discuss and recommend improvements in their work. At General Motors, which has represented both the best and the worst of this approach, thousands of shopfloor discussion groups have proposed changes in everything from wall colors to engineering designs.

QWL efforts have tended to remain limited in scope and to lose momentum after a few years of creative enthusiasm. But far from retreating, the current trend among large companies is towards increased scope of worker involvement. QWL is being followed by a movement involving a more substantial shift, the use of autonomous teams. Here groups of workers take responsibility for scheduling, discipline, and other managerial functions, often eliminating first-line supervisors. Again, the auto industry, the birthplace of the dehumanized assembly line, has been a leader in exploring new alternatives. A new Volvo plant in Uddevalla, Sweden, far surpassing Volvo's earlier autonomous assembly teams, has totally eliminated the line. Small egalitarian, independent groups put together the entire car in a process reminiscent of nineteenth-century craft shops.

The spread of these reforms is impressive. Recent surveys have found that over 50 percent of large companies are using one or another form of employee involvement.

Some writers, such as James Rinehart and Mike Parker, see this whole movement as an effort to undermine organized labor. But while team-based worker participation may discourage unionization, that effect cannot account for the spread of team systems to many industries, such as high-tech, where wages are often high and the threat of unionization is distant.

Nor can these participation efforts be explained merely as disguised speedups. In some instances, simple managerial cynicism and manipulation are evident. But, more often, workers gain real, though limited, powers and report significantly improved working conditions. Almost all studies have also reported increased worker satisfaction and sense of dignity.

To be sure, this movement has not reached all forms of work. Backroom financial services jobs, data entry clerks, fast-food workers, mailroom workers in catalogue houses, low-skill assembly jobs, to name just a few, remain highly routinized, with little genuine worker participation. Yet in a generation many of these may be automated out of existence.

Looking at the economy as a whole, a real transformation of much work organization is under way. Below a bewildering array of programs, occurring in multiple industries and for varying motivations, a consistent movement has been taking shape for three decades. The Scientific Management of work -- breaking it into its smallest components for better control -- has been in constant retreat, and the responsibility of workers has steadily expanded. At several junctures the movement has slowed as it encountered resistance, but it has reemerged each time in a more ambitious form. The current growth of autonomous teams is only the latest example of this advance.

In unexpected ways, developments in the managerial ranks themselves have contributed to reductions in bureaucratic hierarchy. In the golden age of bureaucracy committees and task forces were a relative rarity. The conventional wisdom was contained in the joke that a camel is a horse designed by a committee. Today committees are an accepted part of corporate life, and their scope is growing. Most organizations I studied had recently expanded the use of coordinating task forces. As a result, middle management jobs no longer correspond well to the bureaucratic ideal. Rather than answering to a single boss and focusing on a clearly defined slice of the organization, managers often deal with many different bosses in a shifting alignment of task forces. The scope for initiative has grown as a result, and the dominance of the hierarchy has diminished.

Another, more recent change has been the reduction of management layers. Management layoffs and downswings, previously rare, jumped dramatically starting about 1980. One widely quoted estimate is that 500,000 managers at more than 300 large corporations lost their jobs or were eased out just from 1984 to 1986. Cuts of 20 to 25 percent have been frequent, with an accompanying reduction in the number of layers of authority.

A necessary consequence of this "delayering," as it is now called, has been at least some reduction in rules and procedures. There are simply not enough managers around to enforce them. The cutbacks have thus given a further push to the spread of employee involvement on the shopfloor. And although many suffer from the downsizings, the human consequences on remaining employees seem to be positive. Most workers at all levels seem to appreciate the increased responsibility and discretion.

This complex of changes modifying the old bureaucratic hierarchy is driven not by short-term fads, but rather by long-running forces. Bureaucracy is ill-suited to the demands of advanced markets for innovation and quality. It depends on stable routine, and every significant change requires a reordering of the entire structure. Nor is bureaucracy good at freewheeling invention. The lack of discretion and initiative intentionally built into a tightly controlled bureaucratic company work against both creativity and high quality.

Social shifts have meanwhile also reduced the need for bureaucracy. Scientific Management was effective in controlling a poorly educated work force new to the discipline of industrialization. But now that workers have long become accustomed to the routines of schools and industrial work, the discipline can more easily be relaxed. In an advanced society, managers can better "trust" workers to use discretion in the service of industry.

Thus there is a solid foundation to the corporate search for an alternative to bureaucracy. I do not mean to exaggerate the current rate of change. Only a tiny fraction have instituted widespread autonomous team structures, and even these have not fundamentally challenged the overall hierarchy. But looking over the whole landscape, it is plain that something big is happening. The important question for the long run, then, is what these changes portend. I see four radically different possibilities.

Images of the Future

Underlying the welter of experiments and buzzwords are four contrasting visions of organization intended by corporate leaders, consultants, and critics of contemporary management to address the weaknesses of bureaucracy. These visions have differing rationales and political implications, though, curiously enough, each makes use of the "post-bureaucratic" language of involvement, autonomy, and creativity. Yet the real meaning of these concepts and their long-term potential vary widely.

1. Cleaning Up the Bureaucracy

The first vision of "de-bureaucratization" is deceptive. Although its advocates use the language of change, they really mean to clean up the bureaucracy and return to its pure form.

In the old-line companies -- the General Motors, the AT&Ts, the Dow Chemicals -- leaders frequently talk of "empowerment" and "autonomy." Their actions, however, would look familiar to managers of the 1930s. Above all, these companies want to clarify their business strategy, often engaging in extensive top-level efforts to "define the mission." Then they seek to delineate the responsibilities of each level of the organization to reduce the amount of checking across levels. They establish goals for employees and evaluate them on the attainment of the goals rather than on their adherence to procedures. They analyze spans of control, eliminating unnecessary positions, often entire management layers. In short, they seek to make the organization simpler and more rational.

"Empowerment" and "autonomy," in this context, mean returning to individuals the "power" to perform the roles that have been established for them, allowing them to use their knowledge to carry out the tasks defined by their superiors. This effort is nothing other than a return to true bureaucracy, restoring the effective organization of expertise at its core.

Since it may give more discretion and dignity to some individuals, this effort is, in a sense, a form of participation. I was surprised at first at how positive workers, both blue-collar and managerial, felt about recent cuts in middle management. But it should scarcely be surprising that those who remain feel more challenged and responsible than ever before.

This development, however, rather than being an event of long-run significance, is probably no more than a turn in the life-cycle of bureaucracy. Bureaucratic organizations tend to degenerate over time. The original form -- as developed, for example, by Alfred Sloan of General Motors during the 1920s -- clearly spells out roles and allows a lot of autonomy. But soon mistakes are made, controls are added to prevent a repetition, responsibility is drawn upward, and the top managers get embroiled in details and lose their strategic focus. Meanwhile, managers seek to secure their positions by building personal empires inside the bureaucracy. The end result is a bloated, rule-bound, and conservative organization.

A clean-up of the organization may cut away some of the encrusted fat, but it seems likely only to start a new cycle of degeneration. Layers of management that have been removed begin to be restored. Some managers will say they cannot develop competent replacements when the jump between levels is too great; others will say that motivation is threatened by the scarcity of promotion opportunities; others will simply argue that the quantity of work requires more bodies. Rules start to reappear because, it turns out, different parts of the organization need more uniformity in the way things are done to be able to rely on each other. And, meanwhile, "participation" fades slowly away.

But this approach does not really get to the root problems of flexibility and quality. So other visions of change, more cutting in their implications, have moved to the forefront.

2. Corporations As Closed Communities

The second "post-bureaucratic" image, often mixed with the first, is a workplace community of shared values and commitment. The popular management trend that draws on this image is often called "corporate culture reform." Influenced by the success of Japanese companies as well as some of the American firms profiled by Peters and Waterman, "culture" reformers during the 1980s have emphasized the key importance of strengthening shared loyalties and commitments.

Xerox, for instance, attempted to create a culture of quality through a massive educational campaign. After its top executives met in 1983 to formulate a definition of quality, a new corporate quality office developed the ideas into a 92-page "Green Book," which became the focus of discussions at all levels of the corporation.

Other firms, such as IBM, have long placed great importance on shared values. Among the results at IBM has been an extraordinary commitment to employment security that has lasted to the present day; a strong system of due process and employee rights; and a light touch in the use of formal authority. This kind of corporate commitment to (non-economic) values is, of course, ordinarily what we mean by "paternalism." Its great attraction is that it provides an anchor against the tides of economic opportunism that sweep away so many companies. To create the desired sense of unity, the company must be willing to show concern for members of the community; thus "value-based" or "value-driven" companies, as they are called, tend to have strong employment security. They also create a measure of "participation" by substituting for tight control the voluntary commitment of employees to the good of the whole.

The benefits are clear in both organizational effectiveness and humane concern for employee security. But there are equally important problems. Shared values create strong social boundaries; they mark off those who belong from those who do not. Companies that emphasize distinctive values, therefore, build walls around themselves, and stand in danger of excessive inwardness. They find it hard to bring in outsiders, often turning down job candidates who have records of achievement elsewhere. The inwardness can also produce excessive caution and rigidity among those tied to the firm.

Values are hard to change, even when the environment demands change; they are the most stable aspect of social structure. The effective value-driven companies have been that way since their formation, and they continue to draw on the visions of their original leaders. The effort in recent years deliberately to alter corporate culture through the development of "vision statements" seems (though it is too early to be sure) to be foundering because of the complexity of the task.

AT&T exemplifies the conservatism of corporate cultures. The focus on preeminent technological sophistication and service, drilled deeply into all employees, was a source of strength for many years. Now, however, it blocks attention to the cost control and customer responsiveness needed in the newly competitive environment. Though management is strongly aware of the need for change, the spirit resists the commands of the corporate body.

The focus on enhancing some values paradoxically also raises concerns about values threatened. Paternalist companies invade the privacy of the individual more than purer bureaucracies do. One of the ethical benefits of bureaucracy is that it allows employees to separate their personalities from their jobs and to establish limits on their commitment. Corporate attention to values creates more intense demands. In certain cases, indeed, companies can come to resemble cults.

These concerns are heightened where, as in most cases, a strong hierarchy remains. Although it may be possible to achieve shared values through open and participatory processes, that is not how it has happened so far. The role of charismatic founders remains crucial, and recent attempts at deliberate cultural modification have been orchestrated from the top.

This variation on bureaucracy, in short, does not entail a move away from hierarchy. The Catholic Church is perhaps the purest model of a value-driven organization, but there is no doubt about who is in charge at every level. The same can be said of IBM. Developing a sense of shared commitment in no way conflicts with or diminishes the continued control of strong leaders.

3. Corporations As Free Markets

The preceding two images, I have suggested, represent efforts to modify hierarchy, not to transform it. The third image is more novel. For some companies, life after bureaucracy appears to mean dissolving the old structure of authority and substituting a web of market contracts.

The increased prestige of the market during the Reagan years has unleashed a flood of theorists who want to apply its principles to all spheres of social action. Some corporate reforms, aimed at bringing market criteria inside the corporation, reflect this perspective. Pay for individual performance, the promotion of internal competition, the conversion of departments into profit centers, and the threat -- and practice -- of subcontracting work previously done inside are all examples of this movement.

How does this work in practice? Within a company, each part becomes responsible for meeting the needs of internal "customers" (that is, other departments). Those customers are given the option of going to other competitive sources either inside or outside the company. So, for example, the legal department may no longer be guaranteed responsibility for handling all legal problems; it has to bid against outside firms to get "business" from the company's own departments. A department, or individual, that does a good job is supposed to get more business from within the company. Compensation may no longer depend on formal organizational position and instead becomes linked closely to performance. Loyalty is no longer expected or desired. The basic principle, in other words, is to achieve coordination through market or market-like exchange rather than by a hierarchy of command or by a shared corporate culture.

The fullest flowering of the concept is the growing tendency to subcontract work that was once part of the organization. Many staff jobs, and not a few production jobs, are now spun off into separate firms that sell their services piecemeal to the original corporation. Indeed, many managers who have been laid off in the past decade have "reappeared" in the form of contractors to their old firms.

Among large corporations, General Electric under the leadership of Jack Welch has probably pushed this vision farthest. The job of a leader," Welch says, "is to take the available resources -- human and financial -- and allocate them rigorously." He has publicly declared that loyalty is not a virtue in the company, and he has tried to transform the organization into a set of entrepreneurial units competing for resources.

There is a sense of "participation" in this model as well. It means a liberation of the individual from detailed tasks and the encouragement of innovation, entrepreneurship, and risk-taking. It is, in principle, extremely non-hierarchical since position means nothing next to performance. Employees are expected not to obey but to innovate in meeting the needs of their customers. Welch is one of the most fervent apostles of "empowering" employees: "The idea of liberation and empowerment of our work force," he says, "is not enlightenment -- it's a competitive necessity."

This image is particularly attractive to those of an entrepreneurial spirit and to young employees who are impatient with slow progress through traditional bureaucracies. The "market" model makes room for independent judgment. Yet as a model of organization, the market metaphor is, on reflection, untenable. A corporation cannot really be a pure market. The distinctive capacity of organization is strategic coordination. A market, by definition and by experience, is incapable of strategic planning.

What is possible in a corporation -- and what actually results from this vision -- is a kind of "mixed economy" where a hierarchy sets the parameters for an employee "market." This is where many corporations seem to be moving. The top levels set strategy, determining the customers and basic objectives and setting general goals for their subordinates. Then they get out of the way while competition takes its toll.

This vision seems in the long run to provide none of the human benefits of the market, while creating high costs. It generates highly individualized pressure with little social support; for most people, the pressure becomes wearing after the initial challenge. Furthermore, it paradoxically increases the distance and power of bosses, for they no longer have personal responsibility for the development of their subordinates. They act like impersonal arbiters, setting up the game and judging the winners. Here, for example, is an employee's description of General Electric:

"GE is a very difficult place for any middle manager to operate. ... It's very entrepreneurial, and you can get huge amounts of resources if upper management supports you. It's exciting if you're a risk taker. On the other hand, the power base can be taken away from you at any moment."

In effect, the market model "hides" the power structure of an organization and makes it harder to penetrate. Upper management becomes further removed from day-to-day operations, but its impact is not diminished. By setting up the rules of the game it controls the outcomes.

Will the market model, in this mixed form, expand despite these hidden human costs? There is almost no solid evidence yet of its power or effectiveness, but there are reasons to doubt it. Firms that have come closest to this approach -- including, for example, much of the investment banking and high-tech industries -- often have considerable problems with stability. There is a strong tendency to short-run, opportunistic thinking. The deliberate rejection of loyalty means that employees are quick to jump to other firms for more pay.

They may also have difficulty with internal coordination. This is hard to prove, because in the 1980s' environment of rapid change and growth "freewheeling" forms were often successful. But there is evidence that over time internal competitiveness undermines the cooperation needed to accomplish complex tasks. As the manager quoted above suggests, "politics" in a system like GE's tend to become fierce and destructive. Even investment banks have now begun to see a need for more unity and structure. It should also be remembered, as the economist Oliver Williamson has shown, that hierarchies originally replaced markets because of the inefficiencies of pure contractual relationships in managing large-scale production.

Organizations do, in short, need a sense of unity, a feeling that everyone is in this together, in order to coordinate: the "rigorous allocation of resources" by higher levels is insufficient. The exhilaration of market freedom must be tempered by shared commitments.

4. Corporations As Associations

The final image, like the last one, is of a non-hierarchical system, but one of cooperative association rather than competition. The corporate reflection of this vision is the new stress on "teamwork" and use of consensus-based task forces and committees that some critics have elaborated into models of "ad-hocracy."

Cooperative associations resemble markets in many respects. Both are, in principle, forms of egalitarian relations among free individuals. They share many of the same strengths and weaknesses. Like markets, associations are extremely flexible in dealing with changing demands. They are open at the boundaries and relatively untroubled by the demands of security and loyalty. Both are also strong in invention. Since the parts are not tightly linked, they are free to engage in serendipitous exploration and experimentation.

The differences, however, are also great. A cooperative system encourages relations of diffuse trust rather than of focused contract. It assumes longer-term relationships than the economic model. Personal support and solidarity are valued. And decision-making is an open process of discussion aiming at consensus, rather than an outcome of impersonal market forces.

This is, of the four, the image that comes closest to the ideal of egalitarian participation. In a sense, it would be the "best case" scenario for the future. Cooperative networks suggest the possibility that people acting freely can work together effectively and creatively for collective goals. That sounds good both for the individuals and for the businesses that employ them. The proponents of ad-hocracy who develop these themes sound a seductive note, and some have played it to large crowds.

Several companies in my study are clearly working from this cooperative vision. One medium-sized manufacturing company, for instance, uses consensus-based teams, often crossing multiple levels, for virtually all decisions. Only two levels of management lie between the CEO and the shopfloor, and there are no first-line supervisors. Workers argue with engineers about production processes and often win; they also sit on strategy task forces.

Both blue- and white-collar workers stressed the honesty and openness of the managerial process and the extraordinary personal growth that resulted from this high level of involvement:

"At [another company where I worked] they had teams, but there were always hidden agendas: if we didn't do it right management came back and had to do it again. Here we really analyze the problems together."

"There's growth everywhere: I've really seen people blossom. That's the biggest thing out of it for me."

"This organization is unique in the ability to face product problems with honesty. People will say the systems are not where they should be. People will say we need your help."

The human benefits of such an approach are obvious. The ideal of association is deeply embedded in American values, even antedating the popularity of markets. But the organizational picture is more complex: Attractive as it is, the model does not work very well on a large scale.

Association should, in theory, compensate for some of the key weaknesses of markets. It can achieve strategic coherence through consensus-building. And although loyalty in the strict sense is not required, the trust built into cooperative relations tends to prevent the opportunism that is so dangerous in highly market-based firms. But in practice, trust and consensus are difficult to achieve, and the difficulty grows geometrically with size. From the same company:

"The overall system integration is a real problem. We're very competent in small local groups. But people don't want to leave their home teams and share with other teams."

"I have trouble finding people to make decisions. In theory it allows you to get in touch with the actual end-users and get people with ownership involved in decision making processes.

In actuality often someone has responsibility but not authority; so the decision doesn't stick, or more people want to get in later."

These comments touch on some general problems with cooperative networks. They are generally expensive. A bureaucracy can trim its resources down to the bare bones by focusing sharply on goals, but a network lacks such a clear point of reference. Consensus processes can be clumsy and time-consuming, especially as they move beyond the scope of a small group. Strategic unity is in practice hard to maintain. There is a tendency to avoid major change for fear of the conflict it will cause. While a bureaucratic leader can declare a strategic "revolution," clearing the decks and refocusing, an egalitarian network has a much harder time. When crisis moments do come, an "ad-hocracy" is vulnerable to splits and factions.

Given the difficulties of achieving real consensus, companies have moved again towards a compromise. Task forces and autonomous teams have spread widely in many large companies, but they have not turned relations upside down. For the most part teams are clearly constituted by higher authorities and told what to do, and they remain within a sharply bounded overall organization. The result is, in most cases, a highly layered system; increasing cooperation within layers is balanced by sharp differentiation between them.

In key respects, indeed, many large organizations that have restructured and downsized have increased hierarchy. As layers have been pruned from the middle, the distance between adjacent levels has grown larger; fewer people can make the jump upwards. With the increased need for rapid strategic adaptation, the top has become more active in defining strategic directions, thereby widening its perceived distance from the middle. The formal cooperative institutions -- the autonomous production teams, project groups, and so on -- work within this sharpened hierarchy rather than overshadowing it.

Almost all the companies instituting these associational reforms have very strong, driving CEOs who are also working to centralize strategic authority and to clarify responsibilities. Such organizations are, therefore, simultaneously cleaning up their bureaucracies -- cutting layers, pruning rules, refocusing responsibilities, clarifying spheres of autonomy -- and developing network skills of team building and negotiation. Paradoxically, in some cases such as People Express, the network approach seems necessarily to coexist with a highly charismatic CEO exercising almost mystical and despotic authority. The "team" seems to require an authoritarian "coach."

The Potential Development of Associative Organizations

The search for alternatives to bureaucracy is itself good evidence that the pressures for change are serious. But no one has yet demonstrated a viable alternative that redefines the full hierarchy of a firm. The corporate effort to improve effectiveness and profitability has produced some improvements in work --increases in autonomy and a reduction in the fragmentation of tasks -- but remains limited in scope.

Do these limits mean that bureaucracy is essentially "unshatterable" as Max Weber held? Not necessarily. Of the various types of reform I have described, the one that has made the steadiest progress is cooperative association. After much experimentation, team models have proved themselves at the level of small group discussions. For many purposes, flexible task forces operating consensually seem to work better than individually defined roles. Recently, entire layers of top-down control have been supplanted by teams that cross departments and functions.

Twenty years ago few would have thought that teams of shopfloor workers could successfully manage themselves without supervision; now that is commonplace. Few thought then that task-based organizations could work; now they are everywhere. Theorists of class conflict argued that these models could never overcome management's need for unilateral control. Yet the frontiers of that control have continuously been moved back.

The amount of social invention is already impressive. Our understanding of how to build consensus among diverse groups has been enriched by a flood of new techniques for team building, group problem-solving, and negotiation. A large literature -- generally non-academic and highly oriented to practice -- has grown up around the notion of "process," which essentially means the building of consensus among diverse actors.

On small scales, these skills have by now enabled associative forms effectively to demonstrate their power. Team systems have consistently produced increases in productivity and innovation as well as in worker involvement. But the success of consensual organizations declines rapidly with size and diversity, as problems of coordination grow more serious.

The reason for optimism is that the level of invention remains high. Two recent developments in particular have real promise. First, the rapid advance of computer technology helps complex decision processes by improving communication among dispersed parties and information about alternative courses of action. Second, much excitement and ferment has recently centered on new methods of negotiation --"Getting To Yes," in Roger Fisher's catchy phrase. These techniques have settled some complex problems -- environmental issues, trade relations, community disputes -- that formerly were settled through markets, courts, or administrative fiat. Now they are being adopted by many companies to help coordinate the network of task forces.

It appears, then, that the advance of associative forms will continue. The driving forces that have produced the critique of bureaucracy seem to be gaining rather than losing strength. In the modern economy, bureaucracy no longer enjoys a massive advantage in power and effectiveness over more open and participatory alternatives. But bureaucracy is not yet dead. "Ad-hocracy" still has a lot to learn before it can displace the behemoth.

Postscript

In the background of this essay has been a concern about democracy in general -- political as well as economic. Modern governments, like corporations, have been primarily bureaucratic; the difference has been that the top of the political pyramid has been elected. This political organization has, however, steadily lost legitimacy in the past thirty years: "De-bureaucratization" has become a central theme in the political sphere as in the economic. Governmental structures are increasingly perceived as distant and inflexible, unable to respond to the growing diversity of interest demands. The "Reagan revolution" represents in large part collective despair at the possibility of making the bureaucracy responsive to popular needs -- in effect, a giving up on the current forms of democracy.

Again paralleling the corporation, the market vision has become a powerfully seductive vision for government. Recently, to take one example, the Brookings Institution issued an influential report that argues against democratic processes in public education -- in particular the election of school boards -- on the grounds that they lead to bureaucracy and ineffectiveness. The only solution, in the report's view, is to free schools from politics by subjecting them instead to the discipline of the market; it recommends individual choice through a voucher system.

In the search for an alternative to political bureaucracy, then, the market vision is now a leading contender. But this market vision is antithetical to democratic politics. In setting individuals loose, it denies the importance of agreement and shared direction. It is incapable of building a sense of common aims or public goods.

Mechanisms of cooperative association, however, have been growing at the same time, constituting an alternative solution. With increasing frequency political issues are worked out through direct discussion and negotiation among multiple parties, mediated by government, rather than through the traditional processes of election and administrative rule-making. The decentralizing thrust of the Reagan years has paradoxically contributed to this development. Towns and localities faced with increased authority and pressure for budget decisions have had to seek new methods for resolving disputes. As a result, there has been a burst of innovation in the form of search conferences, negotiated rule-making, and other forms of open discussion.

In school reform, the two models are already in direct competition. Many areas, including the town where I live, are trying to prove that democracy can work better than the market model. Rather than turning individuals loose with vouchers to seek the best school, we can create a structure that enables the members of a community to come together in developing a consensus.

This associational course is clearly not easy. The disagreements are complicated and passions run high. But there have been enough successes in recent years to justify continuing to develop it. It now offers the best possibility for overcoming the limits of our current, bureaucratic form of democracy, and rebuilding a sense of shared political goals.