Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Uncertain~, Diversity, arid Organizational Charlge MICHAEL T. HANNAN There is in our social organizations an institutional inertia.... Unless there is a speeding up of social invention or a slowing down of mechanical invention, grave maladjustments are certain to result. (President's Research Committee on Social Trends, 1933:xxvii.) How difficult is it to reshape complex organizations when conditions change? Ogburn's (1933) work on technical innovation and society built on the premise, illustrated in the quote above, that organizations and social institutions strongly resist change. He argued that the combination of rapid technical innovation and organizational inertia disturbs equilibria. Long periods of disequilibrium caused by lags in adjustment of social structures to changing material conditions can have high social costs, as Ogburn and his collaborators insisted 50 years ago. Despite the seeming ubiquity of organizational inertia in everyday life, the social science literature has sometimes painted a very different picture. Both organizational theorists and specialists in management have often described a world in which organizational adjustment to changing external conditions is almost friction-free. March's (1981:563) review of the liter- ature on organizational change notes this dominant theme: Organizations are continually changing, routinely, easily, and responsively, but change within organizations cannot be arbitrarily controlled.... What most reports on im- plementation indicate . . . is not that organizations are rigid and inflexible, but that they are impressively imaginative. Which is it? Are organizations subject to strong inertial pressures as Ogburn has it? Or do they change easily and routinely as March claims? The work reported here was supported by National Science Foundation Grants SES-8109382 and ISI-828013. 73
74 MICHAEL T. HANNAH This disagreement raises fundamental issues about the relation of organi- zations and society, issues that have important theoretical and practical implications. If change in organizational strategies and structures is rapid and smooth, it is reasonable to respond to changing conditions by continually redesigning existing organizations. But if organizations typically respond slowly or not at all to changing opportunities and threats in their environments, it may make more sense to continually replenish the stock of organizations. These alternative strategies imply vastly different social policies. The disagreement between Ogburn and March reflects more than a gen- erational shift in organizations theory and research. Contemporary opinion among organizational researchers splits sharply on these issues. Questions of organizational inertia are fundamental to understanding organizational structure and change; thus the opposing opinions voiced by Ogburn and March, which continue to divide current researchers, provide a useful frame- work for considering past and present theory and research on organizational change. Inertia is only one of several noteworthy factors affecting the adaptability of organizations to environmental uncertainty. At least as important is the diversity of organizational forms in society the stock of organized solu- tions to problems of producing collective action in variable settings. Trends that eliminate organizational diversity lower the capacity of social systems to deal with uncertain environmental change. Questions about diversity and inertia are especially important when change in technical, social, and political environments is uncertain. If environments are highly stable (and thus certain), there is really no continuing problem of organizational adaptation. It will become clear eventually which forms of organizations are well suited to the stable, prevailing conditions (either by differential selection or by learning and imitation). Likewise, if environments change in predictable ways (for example, seasonal changes in demand for energy, Christmas trees, and other commodities), even highly inflexible organizations can schedule adjustments far enough in advance to match strategy and structures to these changing states. Issues of organizational inertia and organizational diversity are im- portant to understanding modern social change. This essay describes the development of theory and research on organizational processes as it bears on these questions. It also suggests new lines of inquiry that might better clarify the relations between organizational change and large-scale social change. In particular, it discusses recent theory and research that consider organizational diversity and change from ecological and evo- lutionary perspectives.
UNCERTAINTY, DIVERSITY, AND ORGANIZATIONAL CHANGE CENTRALITY OF ORGANIZATIONAL PROCESSES IN LARGE-SCALE SOCIAL CHANGE 75 Most theories in the social sciences emphasize the actions of autonomous individuals, interest groups, social classes, and institutions rather than those of concrete organizations. But almost all modern collective action takes place in organizational contexts; organizations are the main actors in modern society (Coleman, 1982~. When interest groups and social classes take collective action, they do so using specific organizational tools such as labor unions, political parties, or terrorist groups. Recent research shows that even relatively amorphous social protest movements have a higher likelihood of success if they can use existing organizations (filly, 1978~. The state, which has become the focus of so much social action, is itself an organization (or perhaps a hierarchy of organizations). Struggles for power and control in modern societies typically involve struggles between competing organizations for privileged positions in the state structure as well as struggles between the state and other kinds of organizations. Organizations are also important in modern societies because of the role they play in creating, promulgating, and enforcing social norms. The cod- ification of norms as explicit, legitimated organizational rules gives such rules great force. Organizations typically develop formalized roles and procedures for enforcing these rules. Employment contracts, for example, have more continuous and binding effects when labor unions monitor com- pliance and take action to enforce them. Because organizations are key actors in modern society, the speed and direction of large-scale social change are constrained by organizational dynamics. In particular, the responsiveness of society to changing condi- tions depends on the inertia of its constituent organizations and on the diversity of its stock of organizations. The problem of matching outputs of schools to the needs of a changing economy illustrates the problem. It has long been evident that American school systems were failing to teach enough mathematics and science to all but the richest and most able students. Over the past several years a series of national commissions has identified this situation as a "national problem" and urged immediate and far-ranging reforms of U.S. public education. Some commissions urge more attention to teaching (and re- quiring) more mathematics, science, and computing; others emphasize at- tention to writing. All agree that the quality of teachers needs to be upgraded and that more time must be allocated to teaching. A broad consensus seems to have emerged on the definition of the problem; federal and state officials, school district officials, legislators, school employee unions, and parents' groups all urge reform.
76 MICHAEL T. HAXNAN How quickly can the national system of public education be reformed? Despite the fact that many states have imposed new rules and constraints on school systems, there are a number of reasons for suspecting that change in the actual organization of schooling will be halting at best. The demo- graphic and institutional constraints on change in this system are very powerful. Consider the problem of upgrading the technical knowledge of teaching staffs. In a period of declining enrollments, school staffs have been shrinking (although there has recently been an upsurge in demand for science teachers). Given the "last-in, first-out" policies favored by bur- eaucracies and demanded by teachers' unions, change in the composition of teaching staffs will be glacially slow without some radical alteration of employment policies. Any such radical change is sure to encounter stiff resistance from unions, as well as legal challenges. A radical change in policy may also mobilize previously quiescent groups. The complexity of the organizational networks involved compounds the adjustment problem. There is no unitary chain of command; rather there are multiple, partially overlapping jurisdictions of local, state, and federal agencies, with no central planning mechanism. Change in any one sector is hampered by overlaps with others. For example, the seemingly simple problem of changing textbooks in public school systems is made very complicated by the organizational arrangements. Many different organi- zations and individuals must be consulted; any one of them can forestall the change. Implementing even a broad and powerful mandate to change the edu- cational system means changing many organizations and their interlocking connections. The whole system responds only as fast as the slowest com- ponent organizations. Similar issues arise in industry, although the processes are different. In recent years, a number of highly concentrated American industries such as steel, automobiles, and agricultural and construction machinery, have stum- bled before more efficient foreign competitors. The giant American fops in these industries adapted their strategies and structures to earlier technical and social conditions, and they have been ponderous at best in responding to new challenges. Firms in these industries have relied on political muscle to obtain favorable government intervention to limit competition, as we have seen in the auto and steel industries. Success in this tactic serves mainly to further delay radical change in industrial strategies and structures. Global national policies like "reindustrialization" imply massive change in the structures of thousands of organizations. Whether such policies can proceed quicldy enough to meet international competition and rapidly changing technologies depends largely on the responsiveness of existing firms in the economy and on the rate at which new firms can be created and brought
UNCERTAINTY, DIVERSITY, AND ORGANIZATIONAL CHANGE 77 up to speed. Analysis of such policies requires knowledge of the dynamics of organizational populations. The discussion to this point has considered organizations as passive brakes on social changes initiated elsewhere in the society or in the environment. But the image of organizations as passive is seriously misleading. Of course, organizations are constructed as tools for specific kinds of collective action. For example, agents invest resources in hospitals or armies in the hope of achieving specific kinds of performances. But one of the main contributions of organization theory and research has been to show that organizations are far more than simple tools. Organizations consume great quantities of resources in merely maintain- ing their structures. Because great quantities of resources are used for organization building and for bureaucratic or administrative overhead rather than for production or for collective action, organizational politics often revolve around issues of resource allocation (Cyert and March, 1961; Pfef- fer, 19811. Organizational politics makes problematic the relation between technical needs for production and actual distribution of resources. Subunits strive to protect and expand budgets and staff sizes. The resulting com- petition for fixed resources is especially severe in times of contraction or decline (Freeman and Hannan, 1975; Hannan and Freeman, 19781. Because allocations within organizations are subject to intense political contest, organizational action depends on the dynamics of political coalitions. Or- ganizational politics often makes collective action deviate from ostensible goals, from the demands of relevant environments, and from the intentions of organizational leaders. For these various reasons attempts at understanding patterns of large- scale change in modern societies (or relations between public policy and actual implementation) require detailed attention to organizational processes and dynamics. Rational and Natural System Perspectives Systematic organizational theory began when bureaucratic forms gained ascendancy as ways of organizing the activities of the state and of large industrial concerns. German sociologist Max Weber, the founder of socio- logical organizational theory, emphasized the importance of the spread of bureaucracy to the spread of norms of rationality. Bureaucracy, which is built on formalized rules, explicit spheres of competence, and full-time professional staff, permits rapid, efficient, and calculable response to ad- ministrative directives. In Weber's (1978:973) view, The decisive reason for the advance of bureaucratic organization has always been purely technical superiority over any other form of organization. The fully developed bureau
78 MICHAEL T. HANNAN cratic mechanism compares with other organizations exactly as does the machine with the non-mechan~cal modes of production. Because it is precise and efficient and because it can (in principle) serve the interests of any who come to control it, bureaucracy is practically indestructible in Weber's view. Weber's insistence on the machinelike character of modern bureaucratic forms was echoed in this country by Frederick Taylor, the founder of the school of organizational design called Scientific Management (see Perrow, 1979, for a detailed examination of this school). Taylor described smoothly functioning organizations in which all tasks were broken down into minute components according to the logic of "time and motion studies." Research in this tradition sought to learn optimal designs for such organizational machines. Much work in this tradition, for example, tried to discover the optimal "span of control" for industrial organizations, the ratio of super- visors to workers that maximizes efficiency. Much subsequent work in sociology, industrial psychology, industrial en- gineering, and economics was shaped by the broad assumption that organi- zations are efficient, impersonal tools for production, administration, and other forms of collective action. Scott (1981) provides a careful summary and critique of work in this "rational-systems" perspective. This approach has produced much useful knowledge, especially about the conditions under which formal organizations have efficiency advantages in coordinating complex work. Many empirical findings of this tradition have become the conventional wisdom of management and public administration theory. This perspective also continues to guide much current research. For example, an important development in economic theory of organizations argues that organizations are often able to minimize the costs of completing economic transactions when markets fail due to imperfect information, cognitive limitations on the ability to process information, and opportunism (Arrow, 1974; Williamson, 19751. Although the rational-systems perspective continues to shape research on organizations, most sociological research has long made an opposing ar- gument. As early as 1915 German sociologist Robert Michels, who agreed with Weber that bureaucratic forms were indispensable for efficient col- lective action, argued that bureaucracies seldom pursue their ostensible goals. He claimed that organizations are subject to an "iron law of oli- garchy." An organization requires expert leadership even when it is de- signed for democratic and collective ends, as in the case of labor unions and political parties. As leaders learn skills of managing and become dif- ferentiated in prestige and lifestyle from the mass membership, they develop interests in preserving the organization (and their privileged position) at
UNCERTAINTY, DIVERSITY, AND ORGANIZATIONAL CHANGE 79 any cost. They also develop the capacity to control organizational decisions. Thus, Michels argued, leaders typically can and do subvert the goals of the organization to minimize the risk that the organization will be destroyed. According to Michels (1962:364-365), . . . the principal cause of oligarchy in democratic parties is to be found in the technical indispensability of leadership.... Reduced to its most concise expression, the fun- damental sociological law of political parties (the term "political" here being used in its most comprehensive significance) [is] "It is organization which gives birth to the domination of the elected over the electors, of the mandatories over the mandators, of He delegates over the delegators. Who says organizations, says oligarchy." Michels described a process by which an organizational tool takes on a life of its own. One result is that organizational action becomes highly unpre- dictable from knowledge of public goals and interests of its members. This insight has been amplified by numerous studies in the so called "natural- systems" perspective (Scott, 1981), which stresses the continuities between formalorganizations end communities Parsons, 1960;Sel~nick, 19481.Like communities, organizations have rich and complex political systems, and organizational action is often Me outcome of political contests among factions. Subunits of organizations seek to defend self-interests and resist reallocations of resources when conditions change. Moreover, members often develop shared norms in opposition to management. For these reasons organizations are at best "recalcitrant tools," as Selznick (1948) put it. Much early work in the natural-systems perspective involved close ex- amination of the actual process of work in organizations, as in the famous studies from the Hawthorne experiment at the Western Electric works (Roethlisberger and Dickson, 1939~. Also important were the case studies by students of Robert Merton at Columbia such as Peter Blau's (1955) analysis of patterns of exchange in a social work agency and Philip Selz- nick's (1949) study of the relations between the Tennessee Valley Authority and its local community. Recently, organizational sociologists have ex- tended this perspective by conducting comparative quantitative analysis of organizational politics. One particularly useful line of work, which follows the lead of the so-called Carnegie School (especially Cyert and March, 1963), explores how control over essential resources converts to power within organizations and how power balances shape strategy and structure (see especially Pfeffer, 1981; Pfeffer and Salancik, 19781. As in the Weberian tradition, the natural-systems perspective has pro- vided detailed empirical information about the limitations of organizational solutions to problems of collective action. It has identified the processes that distinguish organizations from machines and shifted attention away from idealized images of organizations and toward recurrent patterns of real
80 MICHAEL T. HANNAN organizational action. Numerous findings from this research tradition have also become enshrined in the conventional wisdom of management. Theory and research on organizations during the past 20 years have sought increasingly to synthesize elements of the two starkly opposed perspectives. This work retains the premise of rational-systems theory that organizations are created as tools for collective action and that, in the long run at least, performance matters. That is, this synthetic perspective takes issue with the implicit claim of the natural-systems perspective that organizations are somehow shielded from negative consequences of inferior performance. It also rejects the naive claims of the rational-system perspective that orga- nizations are simple, calculable machines. Instead it treats organizations as open systems Hat depend on a continuing flow of resources from environ- ments. The necessity to maintain such a flow exerts at least some discipline on organizations. However, the fact that one essential resource-member- ship comes with special interests and with attachments to other parts of the social world creates conditions of recalcitrance and inertia. According to various open-system perspectives, organizations are subject both to en- vironmental constraint and to strong inertia. The main theoretical problems concern the relation of these two kinds of constraints. These issues are most interesting theoretically and most relevant to practical problems when they are considered in the context of organizational change. Despite the fact that inertial tendencies seem to be s~ong, especially for old and large organizations, He world of organizations has changed markedly over time. Organizational forms that dominate today differ dramatically from those that held sway a century ago. Chandler (1977) gives a vivid account of the changes in organizational forms in industry over this period. Similar changes can be found in the structures of labor unions, medical care organizations, and government agencies. Thus, changes in social, economic, and political systems apparently do affect organizational structures and practices. The major gaps in our understanding of organizational change concern the actual dynamics- exactly how does change in larger systems affect the distribution of organizationalforms in society? In particular, how much of the change in the organizational world comes about through tinkering (adapt- ing organizational strategies and structures) and how much through replace- ment of one kind of organization by another? We are just beginning to learn about the relative rates of the various processes, which are crucial to an- swering this question. Perspectives on Organizational Change The contemporary literature contains at least three broad perspectives on organizational change. They all emphasize that uncertainty is an inescapable
UNCERTAINTY, DIVERSITY, AND ORGANIZATIONAL CHANGE 81 problem for organizations and plays the key role in shaping their structure and action. The most widespread view, rational-adaptation theory, argues that or- ganizational structures are consciously chosen solutions to certain environ- mental problems. It suggests that the observed variability in the world of organizations reflects planned changes of strategy and structure in response to environmental uncertainties, threats, and opportunities. As a theory of change, this perspective holds that organizations identify threats and op- portunities and reshape structures to mitigate threats and exploit opportu- nities. This approach is mainly directed at explaining the success of large and powerful organizations, those that have managed to adapt well to changing environmental demands. There are numerous variants of this approach, which differ widely in some ways. Contingency theories stress the need for organizations to design structures that buffer their production activities (the so-called technical core) from uncertain environmental variations (Lawrence and Lorsch, 1967; Thompson, 19671. Thus, optimal organizational design is contingent on the nature of the production process and of environmental variations. When either production processes or the pattern of environmental changes shift, organizations attempt to alter their structures, according to this view. In a similar vein, resource-dependence theory argues that organizations must take action to eliminate sources of uncertainty in the environment (Pfeffer and Salancik, 19781. When sources of uncertainty change, organizations are forced to alter their strategies and structures to resolve new threats to their resource flows. An institutional approach, discussed at greater length below, holds that organizational structures are rationally adapted to environmental demands, but that the key demands are often normative and symbolic (DiMaggio and Powell, 1983; Meyer and Scott, 19831. In this view, organizations dem- onstrate their competence within spheres of action and maintain flows of essential resources by displaying appropriate symbols. Such symbolism is often coded in structures. For example, firms display their commitment to planning by creating planning committees or boards of directors and by creating planning departments. What these units actually do is much less important than their mere existence, according to current institutional the- ories. Moreover, as fads and fashions in organizational designs change, organizations are expected to reshape their structures accordingly. As in the cases of contingency theory and resource-dependence theory, the var- iability of structures in the world of organizations is assumed to reflect planned adaptations to changing environmental demands. A second perspective, random-transformation theory, claims that orga- nizations change their structures mainly in response to internal politics and
82 MICHAEL T. HANNAN other endogenous processes, especially the search for solutions to problems of uncertainty. Because there is much randomness in the character of the search, such changes are only loosely coupled with the desires of organi- zational leaders and with the demands and threats of environments (March, 1981; March and Olsen, 1976; Weick, 19761. The third perspective, ecological-evolutionary theory, holds that most of the variability in organizational structures comes about through the creation of new organizations and organizational forms and the replacement of old ones (Aldrich, 1979; Carroll, 1984; Freeman, 1982; Hannan and Freeman, 1977; McKelvey, 1982; Nelson and Winter, 1982; Stinchcombe, 19651. These three perspectives disagree on the sources of organizational di- versity. According to rational-adaptation theory, the diversity of organi- zational forms in society reflects the diversity of environmental problems that must be solved. If the environment becomes more differentiated, di- versity will increase; if it becomes less differentiated, diversity will decline. The random-transformation perspective suggests that diversity reflects mainly the peculiar local and random character of problem solving in each orga- nization. Finally, the ecological-evolutionary perspective states that diver- sity depends on the arrival rate of new organizations and on their diversity, on patterns of environmental vanation, and on competitive dynamics within organizational populations and communities. Progress in explaining organizational diversity and change requires un- derstanding both the nature of organizational change and the degree to which it can be planned and controlled. The remainder of this essay con- centrates mainly on the first issue: does most of the observed diversity in organizational features reflect changes in existing organizations, whether planned or not, or does it reflect changes in populations with relatively inert organizations replacing one another? In other words, does change in major features of organizations over time reflect mainly adaptation or selection and replacement? An Ecological-Evolutionary Approach If organizations are subject to strong inertial pressures and face change- able, uncertain environments, there are strong parallels between change in organizational populations and change in biotic populations. In this case it may be useful to analyze selection and replacement in populations of or- ganizations. As I try to illustrate below, this shift in focus has opened new and interesting questions. A population perspective concentrates on the sources of variability and homogeneity of organizational forms. It considers the rise of new organi- zational forms and the demise of existing ones. In doing so, it pays con
UNCERTAINTY, DlVERSI7'Y, AND ORGANIZATIONAL CHANGE 83 siderable attention to population dynamics, especially the processes of competition among diverse organizations for limited markets. All accepted theories of biotic evolution share the assumption that in- novation, the creation of new strategies and structures, is random with respect to adaptive value. Innovations are not produced because they are useful, they are just produced. If an innovation turns out to have adaptive value, it will be retained and spread through the population with high probability. In this sense, evolution is blind. How can this view be rec- onciled with the fact that human actors devote so much attention to pre- dicting the future and to developing strategies for coping with expected events? Most theorists assume that change in organizational populations is La- marckian, that major changes in the forms of organization come about through learning and imitation. Many kinds of organizations do devote resources to learning and espionage, often seeking to copy the forms of their more successful competitors. In a rough sense, organizations reproduce themselves either by setting up new organizations or by spinning off per- sonnel with the requisite knowledge to copy the form. Nelson and Winter (1982) have developed explicit models of such Lamarckian evolutionary change in populations of business firms. Another line of theory holds that change in evolutionary populations is more Darwinian than Lamarckian (Aldrich, 1979; Hannan and Freeman, 1977, 1984; McKelvey, 19821. This work argues that inertial pressures prevent most organizations from radically changing their strategies and structures once established. It also argues that only the most concrete fea- tures of technique can be easily copied and inserted into ongoing organi- zations. Finally, it emphasizes density-dependent constraints on adaptation by individual organizations: although it may be in the interests of leaders of many organizations to adopt a certain strategy, the capacity of the system to sustain organizations with that strategy is often quite limited. Only a few can succeed in exploiting such a strategy, and "first-movers" have decided advantages. Even when actors strive to cope rationally with their environments, action may be random with respect to adaptation as long as the environments are highly uncertain or the connections between means and ends are not well understood. It is the match between action and environmental outcomes that must be random on the average for Darwinian selection models to apply. In a world of high uncertainty, adaptive efforts by individuals may turn out to be essentially random with respect to future value. The realism of Darwinian mechanisms in organizational populations also turns on the degree to which change in organizational structures can be controlled by leaders. Suppose that individuals learn to anticipate the future
84 MICHAEL T. HANNAN and adapt strategies accordingly, and that organizations simply mirror the intentions of rational leaders. Then organizational adaptations would be largely nonrandom with respect to future states of the environment. On the other hand, if March and others are right, organizational change is largely uncontrolled, and organizations staffed by rational planners may behave essentially randomly with respect to adaptation. In other words, organi- zational outcomes may be decoupled from individual intentions; organi- zations may have lives of their own. In this case it is not enough to ask whether individual humans learn and plan rationally for an uncertain future. One must ask whether organizations as collective actors display the same capacities. The applicability of Darwinian arguments to changes in organizational populations thus depends partly on the tightness of coupling between in- dividual intentions and organizational outcomes. At least two well-known situations generate loose coupling: diversity of interest among members and uncertainty about means-ends connections. When members of orga- nizations have diverse interests, organizational outcomes depend heavily on internal politics, on the balance of power among factions. In such situations collective outcomes cannot easily be matched rationally to chang- ing environments. When the connections between means and ends are uncertain, carefully designed adaptations may have completely unexpected consequences. Moreover, short-run consequences may often differ greatly from long-run consequences. In such cases, it does not seem realistic to assume a high degree of congruence between designs and outcomes. Thus, it may be useful in analyzing patterns of long-term change in organizational forms to supplement Larmarckian theories with Darwinian ones. The fact that members of organizations plan rationally for change and that organizations often develop structures designed to plan and im- plement change does not undercut the value of this view as long as orga- nizations are political coalitions and environmental change tends to be highly uncertain. Organizational Diversity An ecological-evolutionary approach directs attention primarily to or- ganizational diversity. It seeks to answer the question: Why are there so many (or too few) kinds of organizations? Addressing this question means specifying both the sources of increasing diversity, such as the creation of new forms, as well as the sources of decreasing diversity, such as com- petitive exclusion of forms. In other words, an ecology of organizations seeks to understand how social conditions affect the rates at which new
UNCERTAINTY, DIVERSITY, AND ORGAXIZATlONAL CHANGE 85 organizations and new organizational forms arise, the rates at which indi- vidual organizations change structures, and the rates at which organizational populations die out. In addition to focusing on the effects of larger social, economic, and political systems on these rates, an ecology of organizations also emphasizes the dynamics that take place within organizational popu- lations. Questions about the diversity of organizations in society might seem to be of only academic interest. In fact, these issues bear directly on important social issues. Perhaps the most important is the capacity of a society to respond to uncertain future changes. Organizational diversity within any realm of activity such as medical care, microelectronics production, or scientific research constitutes a repository of solutions to the problem of producing certain sets of collective outcomes. These solutions are embedded in organizational structures and strategies. The key aspects of these solutions are usually subtle and complicated. In any large organization, no single individual understands the full range of activities and their interrelations that constitute the organizational solution. Moreover, the subtle aspects of the structure such as "climate" or "culture" defy attempts at formal en- gineering specification. Therefore, it will often prove impossible to resurrect a form of organization once it has ceased to operate. If so, reductions in organizational diversity imply losses of organized information about how to adapt (produce) to changing environments. Having a range of alternative ways to produce certain goods and services is valuable whenever the future is uncertain. A society that retains only a few organizational forms may thrive for a time. But once the environment changes, such a society faces serious problems until existing organizations can be reshaped or new ones created. Since reorganization is costly and may not work at all for the reasons stated above (and because new orga- nizations are fragile), it may take a long time to adapt to the new conditions. A system with greater organizational diversity has a higher probability of having in hand some solution that is satisfactory under changed environ- mental conditions. Adaptation to changing environments in such cases means mainly reallocating resources from one type of existing organization to another. The notion that diversity of organizational forms is a useful hedge against uncertain future changes in environments parallels a classic evolutionary argument. It is, for example, the same kind of argument that has been made against going overboard with the so-called Green Revolution in agriculture. The spread of single strains of crops implies a great reduction in genetic diversity, which may prove problematic if new kinds of pests arise to which the "miracle" crops are vulnerable. Organizational diversity affects society in another way. Since careers are
86 MICHAEL T. HANNAN played out in organizations, the distribution of opportunities for individual achievement depends on the distribution of organizational forms. When diversity is high, individuals with different backgrounds, tastes, and skills are more likely to find organizational affiliations that match their own qualities and interests. For example, the fact that virtually every industry in the United States contains a sizable number of small businesses allows ethnic and immigrant communities to create "ethnic enclaves" within which to develop protected career paths. The presence of such niches in the economy, one kind of organizational diversity, has proven crucial to the economic success of at least some ethnic communities. Diversity is also valued in its own right. Consider the case of the daily press. It is widely agreed in this country that diversity of editorial opinion is a social good and ought not to be sacrificed to economies of business concentration in the industry. Similar views pertain to schooling, higher education, research laboratories, and all sorts of art-producing organiza- tions. How do social, economic, and political environments affect organiza- tional diversity? Almost all attempts to answer this question focus on the controlling role of uncertainty stable and certain environments almost surely generate low levels of diversity. The main theoretical question is: How does environmental uncertainty affect diversity? NICHE THEORY One line of current research in sociology attempts to explain variations in organizational diversity within the context of what population ecologists call niche theory. The concept of niche is used in population ecology to refer to the set of conditions under which some form of life can perpetuate itself. The niche is a mapping between states of the environment and prob- abilities of expansion of numbers. Thus the niche summarizes the environ- mental dependence of a population. Much of the recent progress in bioecological theory has involved embedding naturalistic observations on the structure of niches in nature within Darwinian evolutionary theory (see Roughgarden, 19791. Much work on the evolution of the niche emphasizes niche width. Some forms, called generalists, persist under a very broad range of environmental conditions. Others, called specialists, thrive only in highly specific envi- ronments. Niche theories attempt to explain how patterns of environmental variations affect the evolution of niche widths in biotic communities, that is, how they affect the reproductive success of specialists and generalists. John Freeman and I have argued that many of the classic problems of environmental uncertainty and organizational structure can be recast prof
UNCERTAINTY, DIVERSITY, AND ORGANIZATIONAL CHANGE 87 itably as problems of organizational niche width (Freeman and Hannan, 1983; Hannan and Freeman, 1977~. Organizations clearly vary on this dimension. In our study of American labor unions, we find the Siderog- raphers Union, which seeks to organize a labor force that numbers in the hundreds (siderographers print currency and stock and bond certificates), and the Teamsters, who try to organize almost anyone who works. Likewise, the economy contains firms that produce a single product and others that produce a great range of products. The connection between niche width and diversity is straightforward. To the extent that social trends favor gen- eralist organizations, organizational diversity will decline. But, if specialist organizations have adaptive advantages, the society will contain many di- verse specialists. In other words, the dynamics of organizational niche width constrain organizational diversity. Theories of organizational niche width deal with a "jack-of-all-trades, master-of-none" problem. There are obvious trade-offs between the ca- pacity to withstand a wide range of environmental variations and the capacity for high levels of performance in any one environmental state. In part, these trade-offs concern organizational "slack" or excess capacity. The ability to tolerate diverse environments requires the maintenance of nu- merous routines, patterns of activity that can be invoked by organizational subunits. As Nelson and Winter (1982) convincingly argue, organizations remember by doing, and the capacity to perform a routine declines rapidly with disuse. Generalists, who possess a wide repertoire, must devote con- siderable resources to simply maintaining the readiness of seldom-used routines. Because generalists must commit so many resources to maintaining and rehearsing routines, they sacrifice efficiency and effectiveness in per- forming any single routine. Therefore, at least some specialists usually perform better than generalists in any particular environmental state. Whether there are any adaptive advantages to generalism depends on the rate of environmental change and on the patterns of changes. Levins (1968) proposed a theory of niche width that seems applicable to these organizational questions: Optimal niche width depends on three fac- tors: (1) the magnitude of environmental variations relative to the adaptive capacity of the population, (2) the uncertainty of environmental changes, and (3) the grain of environmental changes. Certainty refers to the odds that the environment will turn up in any particular state; in a maximally uncertain environment each of the possible states is equally likely. Grain refers to the typical durations of environmental states. In fine-grained en- vironments durations are short relative to (unconditional) life expectancy; in coarse-grained environments, typical durations are long. The importance of this distinction is that f~ne-grained environments ought to be experienced (from a selection perspective) as a weighted average of the states. But
88 MICHAEL T. HANNAN populations in coarse-grained environments cannot adapt to some average value; they must have the capacity to withstand long spells in any particular state. Freeman and I formalized the implications of Levins' model for death rates of specialist and generalist organizations in alternative regimes of environmental variations (Freeman and Hannan, 19831. The implications of this model agree in part with the existing literature but differ in one important respect. The conventional wisdom holds that uncertain envi- ronments always favor generalist organizations (see, for example, Katz and Kahn, 1978: 131; Lawrence and Lorsch, 1967:8; Pfeffer and Salancik, 1978; and Thompson, 1967:34-371. The model based on niche theory implies that uncertainty favors generalists only in coarse-grained envi- ronments. We tested this hypothesis using data on the lifetimes of res- taurant firms in 18 California cities. We find that the effect of uncertainty on the relative death rates of specialists and generalists does interact with grain in the predicted way. That is, the niche theory improves on existing theory in explaining the dynamics of niche width in this organizational population. Carroll (1985) has taken a slightly different approach to studying orga- nizational niche width. He points out that the life chances of specialist and generalist organizations depend on the density of each type in the environ- ment. Imagine a market with a center (high, concentrated demand) and a periphery consisting of pockets of heterogeneous demand. In the absence of competition, all organizations in the market concentrate on the center of the market. When the number of organizations competing in the market is high, the largest and most powerful generalists will typically dominate the center. If generalists are numerous, some of them will be forced to exploit more peripheral segments of the market. Because their size and power may allow them to outcompete specialists in the periphery, the life chances of specialists deteriorate when there are many generalists in the market. If, however, one or a few generalists come to dominate and push the other generalists completely out of the market, the opportunities for specialists to thrive in the periphery rise. Thus, concentration in a market should have the opposite effects on the life chances of specialists and generalists. As a market (or organizational field more generally) concentrates, the death rates of generalists will rise and those of specialists will fall. Analysis of death rates in populations of local newspaper firms supports this argument (Car- roll, 19851. Our research group is conducting additional research on the dynamics of organizational niche width among labor unions and semiconductor manu- facturing firms. A number of other groups are working on similar issues using different kinds of organizational populations.
UNCERTAIN, PERSIA, AD ORG^I~TION~ CHANGE 89 INSTITUTIONAL ISOMORPHISM Organizational ecology also speaks to issues of structural i-somorphism- the processes by which organizational structures become matched to features of the environment. Indeed, the initial work on the population ecology of organizations was stimulated partly by Hawley's (1950, 1968) classic ar- gument that organizational structures become structurally isomorphic to those organizations that control flows of resources into a local system. An example of this process is the spread through research universities of plan- ning and budget offices whose internal arrangements are close copies of those of the federal agencies from which universities obtain funding. Hawley's theory was silent on the processes that generate structural isomorphism. Hannan and Freeman (1977) argued that one route to iso- morphism is competition and selection at the population level competitive isomorphism, as DiMaggio and Powell (1983) call it. But clearly this is not the only one. One important new line of sociological theory about organizations iden- tifies institutional processes that produce isomorphism. John Meyer and his collaborator (Meyer, 1978; Meyer and Scott, 1983) have argued that many features of organizational structure are symbols for competencies that may or may not exist. The important adaptive problem for organizations, es- pecially those producing products whose quality is hard to measure, is to evoke the appropriate symbols of competence. Moreover, organization builders are constrained by norms of rationality that dictate a limited number of routines and structures. As general societal processes of rationalization and state expansion pro- ceed, the set of available and endorsed building blocks becomes increasingly homogeneous, and organizational diversity declines. DiMaggio and Powell (1983:147-148) argue as follows: Bureaucratization and other forms of organizational change occur as the result of pro- cesses that make organizations more similar without necessarily making them more efficient . . . highly structured organizational fields provide a context in which individual efforts to deal rationally with uncertainty and constraint often lead, in the aggregate, to homogeneity in structure,. culture, and output. . . . In the initial stages of their life cycle, organizational fields display considerable diversity in approach and form. Once a field becomes well established, however, there is an inexorable push towards ho- mogenization. The argument that general norms of rationality and specific organizational agents (like the state, business schools, and professional associations) create pressures for structural homogeneity, and that these pressures are growing in strength is an important one. But there is a countertendency that must also be considered.
Do MICHAEL T. HANNAN Assume that the population of individuals who demand and use services of organizations is heterogeneous. Assume further that organizational forms partly determine the character of organizational outputs. If, as the institu- tionalists claim, organizations are becoming more homogeneous, the frac- tion of demand for organizational outputs that is either unfilled or dissatisfied with current services will increase. This means that the gains from creating "deviant" organizations to fill this demand will grow. This situation creates opportunities for "outlaw" entrepreneurs to experiment with new organi- zational forms. If any of the experiments are successful, the new forms ought to grow rapidly, lowering the homogeneity of the organizational population. It seems that industrial breakthroughs are often made outside institu- tional channels. The industrial giants in any one era often lack the fore- sight and flexibility to exploit radically new technologies and strategies. Sometimes new modes of production are inconsistent with standard op- erating procedures in the industry, producing conflict that drives creative individuals out of giant firms into entrepreneurial ventures. The U.S. semiconductor industry provides an instructive example (Brittain and Freeman, 1980~. Each of the large vacuum tube producers tried its hand at the production and marketing of semiconductor devices and failed. The industry became dominated by newly created firms. Almost 30 years into the history of the industry, there is still very rapid turnover in the list of leading firms. The crucial organizational innovations that create new industries and new goods and services arise mainly outside the highly institutionalized sphere. In fact, high levels of institutionalization may be a serious im- pediment to innovation. Understanding the forces that create organiza- tional diversity requires analysis of the social forces that shape attempts to create new forms of organizations and of the selection processes that apply to new forms. DISCUSSION An ecological-evolutionary perspective on organizational change rec- onciles the major insights of the two classic traditions of organizational theory and research. It assumes along with rational-systems perspectives that organizations are designed as tools to achieve collective ends. Because organizations compete among themselves for scarce resources, membership, and legitimacy, efficiency in mobilizing each of these affects survival chances. In this sense, organizations face efficiency tests. However, the efficiency testing assumed in current ecological theory is much more complicated than simple testing for technical efficiency in producing some product or service.
UNCERTAINTY, DlVERSlTY, AND ORGANIZATIONAL CHANGE 91 Efficiency in mobilizing resources or in currying political favor may often be more decisive in affecting survival chances than narrow technical effi- ciency. Thus the "rationality" involved in organizational selection pro- cesses may be considerably broader than that envisioned by Weber and Taylor. Nonetheless, ecological-evolutionary perspectives pay explicit at- tention to efficiency testing, broadly defined. Ecological-evolutionary perspectives also build on the natural-systems notion that organizations take on lives of their own. Because organizations must delegate decisions to human actors, they cannot escape processes of political conflict and the creation of subgroup norms and loyalties. Initial patterns of action tend to become enduring bases of political bargaining and of group loyalties. Subsequent attempts to change drastically the struc- ture of an organization encounter both self-interested political objections from subgroups who will lose resources and normative objections to chang- ing rules and structures that have become infused with symbolic value. For these and other reasons, organizations seldom function exactly as planned and are very difficult to reshape. Inertia and change in the composition of organizational populations over time are easy to reconcile within a population perspective. New organi- zations enter many organizational populations at a reasonably high rate. These new entrants are often the carriers of new strategies and structures, the main source of diversity of forms. At the same time, existing organi- zations drop from sight either by simply disbanding or merging with other organizations. The merger process seems to have become the main vehicle by which large and powerful organizations cease to have independent effects on the society. Thus, the populations of business funs, government agen- cies, and others have changed over time in response to the creation of new firms and agencies carrying new forms and following new agendas. They have also changed in response to mergers among existing funs and agen cles. This view of organizational change directs attention to social policies that affect the rate of creation of new organizations. It suggests that dis- cussion of industrial policy, for example, should pay less attention to es- tablished, giant firms than to the social, political, and economic processes that affect the rate at which new firms are started and the life chances of new firms using innovative strategies and structures. More generally, it points to the importance of organizational diversity to society and empha- sizes the need to better understand how social policies affect such diversity. ~ sac ~ em ~ Many of the ideas discussed here were worked out in collaboration with John Freeman. Susan Olzak made helpful comments on an earlier draft.
92 MICHAEL T. HANNAN REFERENCES Aldrich, Howard E. 1979 Organizations and Environments. Englewood Cliffs, N.J.: Prentice-Hall. Arrow, Kenneth J. 1974 The Limits of Organization. New York: W.W. Norton. Blau, Peter M. 1955 The Dynamics of Bureaucracy. Chicago: University of Chicago Press. Brittain, Jack, and John Freeman 1980 Organizational proliferation and density-dependent selection. In John Kimberly and Robert Miles, eds., Organizational Life Cycles. San Francisco: Jossey-Bass. Carroll, Glenn R. 1984 Organizational ecology. Annual Review of Sociology 10:71-93. 1985 - Concentration and specialization: dynamics of niche width in populations of organi- zations. American Journal of Sociology 90:1262-1281. Chandler, Alfred D., Jr. 1977 The Visible Hand: The Managerial Revolution in American Business. Cambridge, Mass.: Belknap. Coleman, James S. 1982 The Asymmetric Society. Syracuse, N.Y.: Syracuse University Press. Cyert, R.M., and J.G. March 1963 A Behavioral Theory of the Firm. Englewood Cliffs, N.J.: Prentice-Hall. DiMaggio, Paul J., and Walter W. Powell 1983 The iron cage revisited: institutional isomorphism and collective rationality in orga- nizational fields. American Sociological Review 48:147-160. Freeman, John 1982 Organizational life cycles and natural selection processes. In Barry M. Staw and Lawrence L. Cummings, eds., Research in Organizational Behavior, Vol. 4. Green- wich, Conn.: JAI Press. Freeman, John, and Michael T. Hannan 1983 1975 Growth and decline processes in organizations. American Sociological Review 40:215 228. Niche width and the dynamics of organizational populations. American Journal of Sociology 88:1116-1145. Hannan, Michael T., and John Freeman 1977 The population ecology of organizations. American Journal of Sociology 82:929-964. 1978 Internal politics of growth and decline. In Marshall Meyer and others, eds., Environ ments and Organizations. San Francisco: Jossey-Bass. 1984 Structural inertia and organizational change. American Sociological Review 49:149 164. Hawley, A. l95o Human Ecology: A Theory of Community Structure. New York: Ronald. 1968 Human ecology. Pp. 328-337 in International Encyclopedia of the Social Sciences. New York: Macmillan. Katz, Daniel, and Robert L. Kahn 1978 Social Psychology of Organizations. 2nd ed. New York: Wiley. Lawrence, Paul, and Jay Lorsch 1967 Organization and Environment: Managing Differentiation and Integration. Cambridge: Harvard University Press.
UNCERTAINTY, DIVERSIlrY, AND ORGANIZATIONAL CHANGE Levins, Richard 1968 Evolution in Changing Environments. Princeton, N.J.: Princeton University Press. March, James G. 1981 Footnotes on organizational change. Administrative Science Quarterly 26:563-597. March, James G., and Johan P. Olsen 1976 Ambiguity and Choice in Organizations. Bergen, Norway: Universitetsforlaget. McKelvey, Bill 1982 Organizational Systernatics. Berkeley: University of California Press. Meyer, John W. 93 1978 Strategies for further research. In Marshall Meyer et al., eds., Environments and Organizations. San Francisco: Jossey-Bass. Meyer, John W., and W. Richard Scott 1983 Organizational Environments: Ritual and Rationality. Beverly Hills, Calif.: Sage. Michels, Robert 1962 Political Parties: A Sociological Study of the Oligarchical Tendencies of Modern Democracy. (Translated by E. Pane and C. Pane). New York: Dover. Nelson, Richard R., and Sidney G. Winter 1982 An Evolutionary Theory of Economic Change. Cambridge, Mass.: Belknap. Ogburn, William F. 1933 The influence of invention and discovery. In the President's Research Committee on Social Trends, Recent Social Trends in the United States. New York: McGraw-Hill. Parsons, Talcott 1960 Structure and Process in Modern Societies. Glencoe, Ill.: Free Press. Perrow, Charles 1979 Complex Organizations: A Critical Essay. 2nd ed. Glenview, Ill.: Scott Foresman. Pfeffer, Jeffrey 1981 Power in Organizations. Marshfield, Mass.: Pitman. Pfeffer, Jeffrey, and Gerald Salancik 1978 The External Control of Organizations: A Resource Dependence Perspective. New York: Harper and Row. President's Research Committee on Social Trends 1933 Recent Social Trends in the United States. New York: McGraw-Hill. Roethlisberger, F.J., and William J. Dickson 1939 Management and the Worker. Cambridge, Mass.: Harvard University Press. Roughgarden, Jonathan 1979 Theory of Population Genetics and Evolutionary Ecology: An Introduction. New York: Macmillan. Scott, W. Richard 1981 Organizations: Rational, Natural, and Open Systems. Englewood Cliffs, N.J.: Pren- tice-Hall. Selznick, Philip 1948 Foundations of the theory of organizations. American Sociological Review 13:24-35. 1949 TVA and the Grass Roots. Berkeley: University of California Press. Stinchcombe, Arthur S. 1965 Social structure and organizations. In James G. March, ea., Handbook of Organi- zations. Chicago: Rand McNally. Thompson, James D. 1967 Organizations in Action. New York: McGraw-Hill. Tilly, Charles 1978 From Mobilization to Revolution. Redding, Mass.: Addison-Wesley.
94 MICHAEL T. HANNAN Weber, Max 1978 Economy and Society: An Outline of interpretive Sociology. 2 vole. Berkeley: Uni- versity of California Press. Weick, Karl 1976 Educational organizations as loosely coupled systems. Administrative Science Quar- terly 21:1-19. Williamson, Oliver E. 1975 Markets and Hierarchies. New York: Free Press.
