From the underlife of institutions to formal systems
If one accepted the systems theory view of authority, given its acceptance of established claims to legitimacy, then it was evident that it had no place for power which did not serve organizational goals. Other forms of power could only reside in unruly spaces where the remit of authority did not extend. These forms of power were something done against authority rather than in its name. Power would be found in the gaps and niches that rationalizing systems neglected or created. Rationalizing systems were equivalent to authority; what they colonized was legit- imate and authorized, a priori, by definition, as it emanated from a rationalizing and sovereign center. What is not authority and is not authorized, what is residual and remains obdurate to the will of rationalizations, must be power. Power repre- sents deviations from normalizing rationalization because it is embedded in those spaces that authority does not extend to. From an organizational point of view, power will invariably be related to resistance because it is outside of and other to authority.
Power was ‘discovered’ in systems theories of organizations from the 1950s onwards, although Mary Parker Follett (1924; also Follett 1918) had anticipated these concerns. We can date formal recognition of its discovery within modern organization theory quite explicitly as 1956:
The usual definitions of power are properly applicable to the internal structures of formal organizations. One reason why research workers have seldom regarded actual power in such organizations may be that the classics of bureaucracy have stressed the rational aspects of organizations, with emphasis on authority to the neglect of unau- thorized or illegitimate power. And it was not that long ago that informal organization was ‘discovered’ in bureaucracies. (Thompson 1956b: 290)
There are several aspects of this discovery of power that are worth remarking. First, it is utterly Parsonian and functionalist in its stress on authority. Second, it sets up a contrast between rational authority and irrational power. Power is what happens when rationality is not secure. Power is irrational and it is illegitimate. Hence, power is the other, or opposite, of authority. In fact, it is what occupies that space that authority has not colonized. Power is brute, uncivilized, and dirty, in contrast to authority, which is rational, legitimate and clean. In fact, the distinction between power and authority is making reference to that distinction between the formal and the informal organization that the open systems conception was supposed to resolve. Other researchers were to echo this distinction as they followed in Thompson’s footsteps. Bennis et al. (1958: 144) made a distinction between ‘for- mal’ and ‘informal’ organization. In the formal organization there resides ‘author- ity’, a potential to influence based on position, while in the informal organization there exists power, ‘the actual ability of influence based on a number of factors including, of course, organizational position’. Formal organization is the repository of authority; informal organization, outside of the formal, is the home of power.
Power is what occurs when authority is not enacted or respected. Power occurs when people who should not exercise their will, in the formal scheme of things, are in a position to do so. Note that these definitional binaries do something rather subtle. They strip power away from the formal structure of the organization entirely; its will is authority, and it is only where this will is resisted or opposed or sidestepped in some way that power exists. The formal organization is now con- ceived of no longer as a thing of power but merely as a thing of legitimate author- ity. However, within the psychological literature, Bennis (1978), amongst others, recognized the formal organization as the main cause obstructing people’s psycho- logical growth in organizations. In this context, authority may become irrational and power can be viewed as a positive resource of the organization, in a manner analogous to the uses that Follett had prefigured.
A number of empirical studies of power in organizations make these distinctions quite clear, with the first being the paper by Thompson (1956b) with which this stream of discussion commenced. As well as making some very general and dubi- ous theoretical points, the paper in question also reported an empirical study. The subject organization was two United States Air Force (USAF) bomber wings. The work of the USAF personnel was characterized by highly developed technical requirements in the operational sphere, for both aircrew and ground crew. While the aircrew possessed greater formal authority than the ground crew, the latter were in a highly central position within the workflow of the USAF base, relevant to the more autonomous aircrew. Briefly, what Thompson (1956b) found surprised him; despite their lower status in the authority system he saw that the maintenance engi- neers who worked on the planes had bases of power available to them that they were able to exercise over the flight crew. These bases were, as is necessary in a sys- tems theory, expressed in terms of general attributes of the system. These general attributes were the ‘technical requirements of operations’, being in a ‘centralized’ position within the organization, and being involved in strategic’ communication’ within it. In other words, the engineers controlled the major source of insecurity for the flight crews, the safety of the plane as a flying machine. In general theory terms this substantive instance was to be translated into control of uncertainty, a move in which the work of Crozier (1964) was vital, as we shall see shortly. The air- crew depended upon the ground crew for their survival and safety, which conferred a degree of power on the latter not derived from the formal design of the base rela- tions. Thompson attributed the power of the ground crew to their technical com- petency vis-à-vis the flight security of the planes and the strategic position it accorded them because of the centrality of concerns for the aircrew’s safety.
Other writers confirmed Thompson’s (1956b) view that it was the technical design of tasks and their interdependencies that best explained the operational dis- tribution of power, rather than the formal prescriptions of the organization design. Dubin (1957: 62), for example, noted how some tasks are more essential to the functional interdependence of a system than others, and the way in which some of these may be exclusive to a specific party. Mechanic (1962) built on this argument, extending it to all organizations, saying that such technical knowledge generally might be a base for organization power. In this way, researchers began to differentiate between formally prescribed power and ‘actual’ power, which was also regarded as illegitimate. Thompson puts it quite precisely when he says that researchers ‘have seldom regarded actual power … [but] have stressed the rational aspects of organi- zation to the neglect of unauthorized or illegitimate power’ (1956b: 290).
Crozier’s (1964) study of maintenance workers in a French state-owned tobacco monopoly, whose job was to fix machine breakdowns referred to them by produc- tion workers, proved absolutely essential for the emerging functionalist consensus concerning power. The maintenance workers were marginal in the formal repre- sentation of the organization design compared to the production workers, who were at the technical core of the organization and central to the workflow-centered bureaucracy that characterized the organization.
In practice, however, the story was very different. The production workers were paid on a piece-rate system in a bureaucracy designed on scientific management principles. Most workers were effectively ‘deskilled’ in a bureaucracy which was a highly formal, highly prescribed organization, where there was little that was not planned and regulated, except for the propensity of the machines to break down, and thus diminish the bonus that the production workers could earn. Hence, to maintain their earnings the production workers needed the machines to function, which made them extraordinarily dependent on the maintenance workers. Without their exper- tise, breakdowns could not be rectified or bonus rates protected. Consequently, the maintenance workers had a high degree of power over the other workers in the bureaucracy because they controlled the remaining source of uncertainty. Management and the production workers were aware of this situation and had attempted to remedy it through preventive maintenance. But manuals disappeared and sabotage sometimes occurred. The maintenance workers were indefatigable in defense of their relative autonomy, privilege and power. Through a skilled capacity, the result of their technical knowledge, they could render the uncertain certain. The price of restoring normalcy was a degree of autonomy and relative power, enjoyed and defended by the maintenance workers, well in excess of that formally designed within a system whose limits they defined. Workers knew well how to exert their knowledge to control the speed of the line as well as how to stop it. Soldiering was a practice sustained by the control of knowledge that others – the supervisors – did not have. To control these practices, organizations constantly evolve new systems and procedures but they can never do so fully. There are always new spaces of knowledge controlled by a few. Even outsourcing places the organization in a delicate relation to the contracted outsourcer at contract renegotiation time.
Crozier’s (1964) study was a landmark. He had taken an underexplicated concept – power – and had attached it to the central concept of the emergent theory of the firm, which was uncertainty. In a different way, March and Simon (1958) recognized the role of uncertainty (bounded rationality) in organizations and sug- gested effective structures to confront it. A central feature of organizations as they were conceptualized in the ‘behavioral theory of the firm’ (Cyert and March 1963) was that they attempted to behave as if they were systems. Yet, they did so in an uncertain environment. The ability to control that uncertainty thus represented a potential source of power. Crozier subsequently revisited the links between power and uncertainty as a critical resource (Crozier and Friedberg 1980). Members of an organization meet each other in spaces that offer relatively open opportunities for control of rules and resources. People do not adapt passively to the circumstances that they meet; they use these circumstances creatively to enhance the scope of their own discretion, through shaping and bending rules and colonizing resources. Power was still seen in terms of the control of uncertainty as it was played out in daily struggles over the rules of an uncertain game, largely in zero-sum terms.
After Crozier (1964) and Thompson (1967), the field developed rapidly. A theory emerged, called the ‘strategic contingencies theory of intra-organizational power’ (Hickson et al. 1971), which built on these ideas. It explicitly acknowledged its debt to Thompson’s (1967: 13) ‘newer tradition’ that conceives of the organiza- tion as ‘an open system, indeterminate and faced with uncertainty, but subject to criteria of rationality and hence needing certainty’. The approach combined Crozier’s (1964) stress on ‘uncertainty’ with sociological functionalism and the behavioral theory of the firm in economics (Pennings et al. 1969), as well as Blau’s (1964) social exchange theory, and operationalized a definition of power that used Dahl’s (1957) and Kaplan’s (1964) behaviorist approaches to its definition. We shall return to this aspect in the next section.
The open system had a behavioral essence: Limitation of the autonomy of all its members or parts since all are subject to power from the others; for sub-units, unlike individuals, are not free to make a decision to participate, as March and Simon (1958) puts it, nor to decide whether or not to come together in political relationships. They must. They exist to do so. (Hickson et al. 1971: 217)
The subsystems of the open system organization are conceptualized as if they were autonomous actors, with goals and values that cohere. This is, of course, a theoret- ical abstraction. Their behavior is guided by the economic rationalism of the open- system/bounded-rationality construct, so they are inescapably free in a double voluntarism and determinism, because they are only free to do what they must. Thus, there is no alternative for the subsystem but to be freely subjected by its own natural rationality. Subunits of organizations conceived as systems may be con- ceived as having powers of action but the assumption that these powers are unified and homologous, that they override competing interests and subjectivities, strate- gies and meanings, is a theoretical a priori, not an empirically nuanced observa- tion. To assume a subsystem is to make an assumption of unitary calculability inhering in it. It renders the conceptual abstraction of the organization as system as strangely flat and one-dimensional. Authority is condensed down into the sub- units while the relations between them are conceptualized as inherently ones of power. But the power that is conceptualized is not relational but structural; if one can identify the subunit that handles most uncertainty for the system, with a few caveats one will have identified power, irrespective of what these unitary actors do. In fact, they can do anything, because what they do does not enter into the theory. The theory is purely structural.
The core idea is that these subunit subsystems reduce uncertainty for the overall organization. In doing this they exchange resources with each other and with the environment to which they are open. The institutional environment is the source of change in goals (Pennings et al. 1969: 420); thus, it is a key factor introducing uncertainty into the rational system. To the extent that organizations are more structurally dependent on certain subsystems to cope with the uncertainties gener- ated from the environment, from the processing of those inputs that the subunits receive, and from their output, then these subunits, the theory predicts, will be most powerful. Power is thus related to control of uncertainty, as this is measured by responses to a formal survey, in which departmental managers, who were pre- sented with a series of hypothetical scenarios for evaluation, responded in terms of Likert scales. In this way, the functionally specific subunits that used esoteric tech- nical knowledge to control uncertainty and thus increase their power relative to the formally designed hierarchy were identified.
The four subunits, not surprisingly, seemed to correspond to Parsons’ four func-tional imperatives. The subunits were interdependent, but some were more or less dependent, and produced more or less uncertainty for others. What connected them in the model was the major task element of the organization, ‘coping with uncertainty’. The theory ascribed the balance of power between the subunits to imbalances in how these interdependent subunits coped with this uncertainty. Thus the system of subunits was opened up to environmental inputs, which repre- sented a source of uncertainty. Subunits were characterized as more or less special- ized and differentiated by the functional division of labor, and were related by an essential need to reduce uncertainty and achieve organizational goals ‘to use dif- ferential power to function within the system rather than to destroy it’ (Hickson et al. 1971: 217).
According to this model, power is defined in terms of ‘strategic contingency’; in parallel and related work it was to be defined as ‘strategic choice’ (Child 1972). Strategically contingent subunits are the most powerful, because they are the least dependent on other subunits and can cope with the greatest systemic uncertainty, given that the subunit is central to the organization system and not easily substi- tutable. The theory assumes that the subunits are unitary and cohesive in nature whereas, in fact, they are more likely to be hierarchical, with a more or less prob- lematic culture of consent or dissent. To be unitary, some internal mechanisms of power must exist to allow such a representation to flourish, silence conflicting voices, and overrule different conceptions of interests, attachments, strategies and meanings. The theory assumes that management definitions prevail but research suggests this is not always the case (Collinson 1994). Nor can we assume that manage- ment itself will necessarily be a unitary or cohesive category. For management to speak with one voice usually means that other voices have been marginalized or silenced.
For a long time the key journals found space for contingency contributions that provided very little insight into important aspects of power. Even with the mani- festation of a strategic contingencies theory of power, existing patterns of legitimacy were not challenged because hierarchy was not addressed. If anything, in contem- porary systems theory, the situation has deteriorated further. While, recently, open systems theory has been revivified by diverse developments in chaos and complex- ity theory, these newer approaches to systems, which stress how organizations cope with rapidly changing environments through complex evolutionary adaptations, tend to downplay the role of agency and thus power (Kaufmann 1993). Change is increasingly seen in terms of complex and somewhat random interactions between large numbers of elements in interaction in computer models (Sorenson 2005).
Source: Clegg Stewart, Courpasson David, Phillips Nelson X. (2006), Power and Organizations, SAGE Publications Ltd; 1st edition.