A manager signs off on a project, but the engineer who controls an essential system can determine what options are actually available. A well-connected coordinator can bring people together or keep them apart. Each person may have influence, yet the sources differ. Organizational power is the capacity to affect decisions, resources or others’ behavior within an organization. It can come from formal authority, control of scarce resources, knowledge, network position and dependence relationships. Its study is more precise when it asks what decision is at stake and who can change another actor’s options.
Power is often confused with status. Status is the respect, esteem or standing granted by others; power concerns control over valued outcomes. A person may be admired without controlling a budget, or hold a powerful office without being admired. Magee and Galinsky’s review treats these as distinct bases of hierarchy that can reinforce one another. The distinction matters because an intervention directed at recognition may not change decision rights, while a change in formal authority may leave informal standing intact. A useful analysis measures both rather than assuming rank means the same thing in every interaction.
Where power comes from
Astley and Sachdeva organized structural sources of intraorganizational power around hierarchical authority, resource control and network centrality. A hierarchy grants a role formal rights to decide. Resource control matters when others need money, information, staffing or access. Network centrality matters when someone connects groups that would otherwise struggle to coordinate. These are a theoretical synthesis, not a fixed list of every possible source. Their strength lies in making power observable through relationships and task dependencies rather than treating it as a personality trait.
Formal authority is the legitimate right attached to a position. It can specify who approves spending, sets priorities or evaluates performance. Yet a written rule does not ensure that the named person determines the final choice. Aghion and Tirole distinguished formal authority from real authority, the effective control over decisions. Information, urgency, overload and delegation can shift real authority toward someone with expertise or access while the formal leader retains a signature. An organization that wants to understand power must examine who frames alternatives and who can veto them in practice.
Expertise can grant influence because the decision maker depends on information the expert holds. That dependence can be beneficial: an organization should often defer to specialist knowledge. It can also create opacity if expertise is treated as a reason to prevent scrutiny. The distinction is not whether an expert “has power” in the abstract; it is which decisions would become difficult without their cooperation, how alternatives are evaluated, and what accountability applies. A transparent process can preserve expert input while preventing one person from controlling every interpretation of the evidence.
Resource control works similarly. A team with a scarce tool, a critical dataset or a distribution channel can negotiate from a stronger position because others need access. The source of power may change when substitutes emerge. If another team develops a compatible tool, dependence falls even though titles remain the same. This relational character is easy to miss in an organization chart. A study of power should identify what is valued, who controls it, how substitutable it is and whether the dependency is mutual.
Network position can matter when someone sits between groups that exchange information. A coordinator may not have the most senior title, yet can influence which problems reach leadership and which collaborators meet. Research on potential power and power use found that hierarchy and network centrality, along with influence behavior, related to others’ perceptions of a person’s power. This is not proof that every central person manipulates information; it shows that position and behavior can jointly matter. A network map must also consider whether a tie actually carries consequential information.
Potential power and its use
A person can possess capacity without using it in a particular episode. A supervisor may be able to block a proposal but choose not to. An expert may be consulted yet decline to steer a decision. Distinguishing potential power from exercised power prevents inference from title alone. The same formal structure can support different behaviors across projects. Researchers can examine meeting records, resource allocations, decision histories and participant accounts to understand how capacity became action. No single measure reveals the full pathway from opportunity to outcome.
Use of power can be cooperative or coercive. A manager may remove an obstacle so a team can work, or use approval rights to punish dissent. The physical act of deciding may look similar, but the effect on information flow and options differs. A rigorous account therefore specifies the target, means, constraints and consequence rather than labeling power intrinsically good or bad. Governance should make consequential decisions reviewable and provide channels for appeal, especially when affected people cannot easily exit the relationship.
Power may change the psychology of the person who holds it. Keltner, Gruenfeld and Anderson proposed an approach-and-inhibition account: greater power can reduce constraints and orient attention toward rewards, while lower power can increase attention to threat and others’ requirements. This is a theory and evidence synthesis, not a rule that every manager becomes impulsive or every subordinate fearful. Galinsky and colleagues’ experiments found that manipulated power reduced some forms of situational influence under controlled tasks. Generalizing those effects to a specific organization requires observing its norms, incentives and accountability.
The same flexibility that allows decisive action can also hide mistakes. Someone who feels able to act without checking others’ interpretations may overlook information held by people with less authority. But powerful actors can be attentive, and low-power actors can shape outcomes through coalitions or expertise. An article should avoid using a group average as a personality diagnosis. Better questions concern whether decisions invite challenge, whether alternatives are recorded, and whether those affected can raise concerns without retaliation.
Anderson and Brion reviewed power as a dynamic process involving acquisition, maintenance and loss. A person may gain power through useful contributions or alliances, retain it by controlling information, and lose it when dependencies shift or legitimacy weakens. This temporal view matters because a single snapshot cannot tell why someone is influential. It also complicates a simple claim that power always perpetuates itself. Some forms create advantages, while changing tasks and relationships can unsettle them. Research should follow the pathway rather than infer destiny from present rank.
Power is relational
An individual’s power depends partly on others’ alternatives. If a worker has several employers, one manager’s ability to impose conditions may be lower than when leaving is costly. If a department can obtain data elsewhere, another department’s control weakens. These examples express a dependence logic rather than a claim that power can be measured by salary alone. They also show why structural changes, such as creating backup capacity or transparent data access, can alter power without changing anyone’s personality.
Multiple parties can hold different kinds of power in the same decision. A regulator can define a rule, a technical team can define feasibility, a union can coordinate workers, and a customer can decide whether to use the result. These relations may be asymmetric yet interdependent. A map of who signs the final document will miss who set the agenda or constrained acceptable alternatives. To understand organizational power, researchers should trace the sequence from problem framing through information gathering, negotiation, decision and implementation.
Power can also be distributed intentionally. Delegating a decision may bring expertise closer to the work and speed response, while central oversight can protect consistency or accountability. Aghion and Tirole’s formal/real authority model helps explain the tradeoff: giving a subordinate effective control can increase initiative but reduce the principal’s direct control. The model identifies a design tension, not a universal answer. The appropriate balance depends on the cost of errors, knowledge distribution, time pressure and rights of those affected.
A team may say that everyone has a voice while one person controls the agenda and budget. Voice is valuable, but speaking is not equivalent to decision power. A fair process should clarify whether input is advisory, binding or subject to a reasoned response. If a request for feedback is merely ceremonial, people can learn that participation does not alter outcomes. This is an Institute synthesis informed by the power literature, not a measured treatment effect. It points to the need to examine how participation connects to actual decisions.
Measuring a moving relation
Surveys can ask who influences decisions, but respondents may not see hidden negotiations or may avoid naming someone powerful. Organization charts reveal formal rights but not real authority. Network analysis can identify central positions but does not show whether a tie carries control over a specific resource. Decision tracing can reveal who proposed, blocked or revised an option, yet records may be incomplete. A robust study combines methods and defines the decision domain. “Who has power here?” is too broad unless the outcome and time frame are specified.
Experiments can isolate psychological effects of temporary power manipulations, while field studies better describe sustained organizational arrangements. Neither alone resolves every question. A laboratory prime is not the same as a year of supervisory responsibility; a field correlation can reflect selection and unmeasured constraints. A careful synthesis lets each design answer its own question. Claims about how power changes behavior should specify whether they concern perceived power, actual control, role assignment, or the consequences of one decision.
Equity matters because dependencies and exit options are unequal. A person whose income, access or reputation depends on a superior may face greater cost for dissent than someone with alternatives. The organization’s formal grievance channel may exist yet be difficult to use in practice. Studying real power means investigating those costs and whether a challenge receives a fair response. It also means acknowledging differences within a nominal category such as “employees.” The relation to a specific resource or decision is more informative than a single hierarchy label.
Practical governance
An organization can make power more visible by documenting decision rights, criteria and avenues for review. It can identify single points of dependence on a person or system and create appropriate backups. It can separate technical advice from final approval when that preserves both expertise and accountability. These are design proposals, not proven universal interventions from the cited studies. They should be evaluated for speed, fairness, errors and unintended burdens in the actual organization. The goal is not to eliminate influence but to make consequential influence accountable.
A leader can also ask what information they are unlikely to hear. If disagreement carries a career cost, silence is not evidence of agreement. A structured invitation is useful only if the subsequent response shows that information may change a decision. In some settings anonymous reporting helps; in others it can undermine dialogue or fail to protect the reporter. No single tactic substitutes for examining the dependence structure. The most useful question is whether a person can present a consequential fact without facing disproportionate risk.
What the evidence does not show
The evidence does not show that formal rank fully determines real power, that network centrality always brings control, or that one psychological effect of power applies to every leader. Structural models, laboratory experiments and field studies illuminate different pieces. A survey perception of power is not identical to a documented ability to alter an outcome. Status and power can correlate, but they remain distinct constructs. Treating all influence as misconduct would also be misleading: legitimate authority and expertise are necessary for many coordinated tasks.
Nor does the evidence identify one ideal level of centralization. Delegation can improve initiative and responsiveness while creating coordination or oversight problems; concentration can support consistency while suppressing information. The balance is a design and governance question informed by the task, affected people and failure costs. Claims that a new structure “empowers everyone” need evidence about actual options, not only an updated chart. A change in rights should be followed through to decisions and lived consequences.
The manager, engineer and coordinator in the opening example hold different capacities to shape a project. Understanding their power requires following the resources, information and choices that connect them. That analysis is more useful than assigning each a single rank on an invisible scale. It helps an organization preserve expertise, share responsibility and check abuse by making the sources and exercise of influence open to scrutiny.
Continue exploring
Research references
- Astley, W. G., & Sachdeva, P. S. (1984). Structural sources of intraorganizational power: A theoretical synthesis. Academy of Management Review, 9(1), 104–113. View source
- Aghion, P., & Tirole, J. (1997). Formal and real authority in organizations. Journal of Political Economy, 105(1), 1–29. View source
- Magee, J. C., & Galinsky, A. D. (2008). Social hierarchy: The self-reinforcing nature of power and status. Academy of Management Annals, 2(1), 351–398. View source
- Brass, D. J., & Burkhardt, M. E. (1993). Potential power and power use: An investigation of structure and behavior. Academy of Management Journal, 36(3), 441–470. View source
- Keltner, D., Gruenfeld, D. H., & Anderson, C. (2003). Power, approach, and inhibition. Psychological Review, 110(2), 265–284. View source
- Galinsky, A. D., et al. (2008). Power reduces the press of the situation. Journal of Personality and Social Psychology, 95(6), 1450–1466. View source
- Anderson, C., & Brion, S. (2014). Perspectives on power in organizations. Annual Review of Organizational Psychology and Organizational Behavior, 1, 67–97. View source