Organisational Power: Definition, Sources and How It Is Exercised

Organisational power is the capacity of one party to influence the behaviour of another so that the second acts in a way they would not otherwise have chosen. The definition is deliberately behavioural: power is identified by its effect, not by a title, a budget line or a position on a chart.

Three features follow from it, and each is regularly lost in summaries.

Power is relational. It exists between parties, not inside a person. A manager with power over a team of graduates may have none over a specialist the organisation cannot replace.

Power is potential. It need not be exercised to exist. A great deal of organisational behaviour is shaped by what people anticipate and not by any instruction actually given.

Power is dependency-based. Whoever controls something another party needs, and cannot easily obtain elsewhere, has power in proportion to that dependency. This is the observation that unifies every taxonomy that follows.

The five bases of power

French and Raven's 1959 taxonomy remains the standard starting point because it separates sources that behave differently.

Coercive power rests on the ability to punish. It produces compliance while observation lasts and generates resistance and concealment when it does not. It is also self-limiting, because using it expends it: an unenforced threat is stronger than an enforced one.

Reward power rests on the ability to give something valued — pay, promotion, assignment, recognition. It works only to the extent the recipient values what is on offer and believes the giver can actually deliver it, which is why reward power evaporates in organisations where promised rewards have failed to arrive.

Legitimate power rests on an accepted right to require something by virtue of position. It is the cheapest form to exercise and the most fragile, because it depends on a shared belief about the role, not on anything the holder controls. Restructurings damage it disproportionately.

Referent power rests on identification: the wish to be associated with or approved of by the holder. It transfers to what the holder asks for without needing justification, and it cannot be conferred by appointment.

Expert power rests on knowledge or skill the other party needs. It is narrow — confined to the domain of the expertise — and it decays as the knowledge diffuses or the problem changes.

French and Raven later added informational power, resting on the content of a message rather than on the person delivering it, and it behaves unlike the others: once the information has been transmitted, the influence persists and the holder has lost the advantage.

Structural sources that outlast personal ones

The five bases describe power held by individuals. Pfeffer's later work, and the strategic contingencies theory of Hickson and colleagues, describe where power accumulates in organisations regardless of who occupies the post.

Dependency and substitutability. A unit whose work others depend on, and which cannot readily be replaced by an external supplier or another department, accumulates power. The corollary is uncomfortable and well evidenced: outsourcing a function reduces the power of whoever runs it, independently of how well they run it.

Coping with uncertainty. Units that absorb uncertainty on behalf of the organisation gain influence disproportionate to their size. Legal and regulatory functions in a regulated industry, safety in an industry with a poor incident record, and treasury when credit tightens all rise for this reason and recede when the uncertainty does.

Centrality. Being positioned so that work flows through you gives access to information and to timing. A scheduling function may be junior and still be consulted on decisions well outside its remit.

Control of resources and of the agenda. Budget is the obvious case. Less obvious, and more durable, is control over what gets discussed: the person who drafts the agenda and the person who writes the minutes exercise a form of power that leaves very little trace.

Power, authority and influence

The three words are used interchangeably in ordinary speech and mean different things here.

Authority is legitimate power attached to a position, conferred by the organisation and bounded by it. It flows downwards and can be withdrawn.

Power is the broader capacity, of which authority is one source. It flows in any direction, including upwards: a personal assistant who controls access to a chief executive holds real power with no authority at all.

Influence is power in use. It is the observable process — persuasion, coalition, appeal to rule, exchange — by which a potential becomes an outcome.

The distinction matters for diagnosis. A manager who complains of being unable to get things done despite having the authority is usually describing a shortage of expert or referent power, and more authority will not supply it.

What people actually do with it

Empirical work on influence tactics is consistent across studies, and the pattern is more useful than the taxonomy.

Rational persuasion, consultation and inspirational appeal are the tactics most often associated with commitment rather than mere compliance. Exchange and ingratiation produce compliance. Pressure and legitimating tactics — pointing to the rule — produce compliance at best and resistance frequently.

Direction of influence matters. Rational persuasion is the dominant upward tactic because subordinates rarely have anything else available. Pressure is almost exclusively downward. Coalition is the characteristic lateral tactic, because peers cannot instruct one another and must build agreement instead.

The practical reading is that tactics chosen for their speed tend to produce the weaker outcome. Pressure gets an action; consultation gets an action plus the willingness to flag a problem with it next time.

How power is acquired, and how it is lost

The bases describe what power rests on. What a student is usually being asked is how someone comes to hold it, and the mechanisms are reasonably well established.

Power accumulates where someone takes on a problem nobody else wants and it turns out to matter. This is the most reliable route in practice and the least discussed, because it looks like willingness and not strategy. A manager who absorbs a regulatory workstream during a quiet period holds real influence when the regulator arrives.

It accumulates through position in the communication network. People who bridge between groups that do not otherwise talk hold information neither group has, and that position is created by where someone sits, not by what they know.

It is lost in three characteristic ways. The expertise diffuses, because the problem becomes routine and others learn to solve it. The uncertainty resolves, so the unit that absorbed it becomes an overhead. Or the dependency is engineered away — by a second supplier, a documented process, a system that captures the knowledge the individual held.

The last of these is why knowledge hoarding is so persistent despite being obviously corrosive. Documenting what you know reduces your own power in the precise sense the model describes, and an organisation that asks for documentation without changing what it rewards is asking people to act against their interest.

Reading power in a live organisation

Three questions get further than an inventory of job titles.

Who is consulted before a decision is announced, as distinct from who signs it? The signature records authority; the consultation records power.

What happens to proposals that are never made? An issue that everyone privately agrees is a problem and nobody raises is the clearest available evidence of the agenda-setting form, and it cannot be seen by asking who won an argument.

Who can be overruled without consequence? The parties whose objections are routinely noted and then passed over are, by definition, the ones without the dependency that would make the objection expensive to ignore.

Why the concept has a bad reputation

Students frequently write about organisational power as though it were a euphemism for manipulation, and the resulting answers score badly because they moralise instead of analysing.

The neutral framing is the accurate one. Every organisation that coordinates the work of more than one person distributes the capacity to influence unevenly, and it has to: a structure in which nobody could get anyone else to do anything would not function. Power in this sense is a property of organisation, not a deviation from it.

What is open to criticism is how it is distributed and how it is used, and those are separate questions from whether it exists. An answer that establishes where power sits, and only then asks whether that distribution serves the organisation's stated purpose, is doing analysis. An answer that treats the existence of political behaviour as the finding has stopped at the first step.

Where the concept is contested

Two criticisms are worth carrying into an assignment.

The first is that the French and Raven categories overlap and are not independent. Reward and coercive power are the same lever in opposite directions, and legitimate power usually carries both with it. Treating them as five separate sources overstates how separable they are in practice.

The second is that the whole framing is individualistic, and deals only with power that is visible in a decision. Lukes' three-dimensional account argues that the more consequential forms operate earlier: in keeping items off the agenda entirely, and in shaping what people believe to be possible so that no conflict arises to be observed. On that account a workforce that never raises an issue may be evidence of power at its most complete, not of its absence — which is precisely what a taxonomy built on observable influence attempts cannot detect.