The Six Bases of Power, and Which One You Actually Have

Two people make the same request in the same meeting. One gets a polite deferral. The other gets agreement within a minute. Neither is the other's manager, and their job titles are roughly equivalent.

This happens often enough that most people have a folk explanation for it, usually involving charisma, or politics, or who is friends with whom. None of those are wrong exactly, but they are too vague to act on.

There is a more precise account, and it has been around since 1959. It says influence comes from six identifiable sources, that they behave very differently, and that most people rely on the weakest one available to them without realising there was a choice.

Why do some people without authority get their way?

Because authority is only one of six sources of influence, and it is not the most durable.

In 1959, social psychologists John French and Bertram Raven published a framework identifying five bases of social power. Raven later added a sixth. It remains one of the most widely used models in the study of organisational behaviour, and its usefulness is that it separates things people usually lump together.

The critical division is between power that comes from your position and power that comes from you. The first is lent to you and can be taken back. The second is yours and travels.

The six bases, and where each one comes from

Legitimate power comes from your formal position. People do what you ask because your role entitles you to ask. It is efficient, it requires no relationship, and it evaporates the moment you are no longer in the role.

Reward power comes from your ability to give people things they want: money, promotion, good assignments, recognition. It works reliably while you control the rewards and stops when you do not, or when the reward stops mattering.

Coercive power comes from your ability to impose costs. It produces the fastest visible compliance of any base and the worst long-term results, for reasons worth examining separately.

Expert power comes from knowing things that matter. People defer because you are more likely to be right. It cannot be revoked by an organisation, it transfers when you leave, and it decays only if the knowledge becomes obsolete.

Referent power comes from people wanting your regard. They are influenced because they respect you, identify with you, or want your good opinion. It is the most durable of the six and the slowest to build.

Informational power comes from controlling or possessing specific information. Unlike expert power, which is about capability, this is about access to particular facts. It is situational and often temporary, since information spreads.

How this is taught inside Astra Trainer

These bases are core vocabulary in the Diplomacy & influence direction of the Leadership, Career & Influence world, fourteen courses on reading a room, earning trust before you ask, and bringing people to yes without pressure or manipulation.

The terms also appear in the world's daily Connections round, where legitimate, referent, expert and coercive show up as a set, so the distinctions get drilled rather than read once and forgotten. That matters here more than with most frameworks, because the six are easy to nod along to and genuinely hard to apply in the moment unless they have become automatic. Lessons take about five minutes and a guide walks you through anything strange.

Which kinds survive you losing your job title?

This is the question that reorganises how you think about the whole framework.

Legitimate, reward and coercive power are all lent. The organisation grants them with the role and withdraws them with it. Someone who has spent a decade building influence entirely on these three discovers on the day they change jobs that they have nothing, and the discovery is usually a surprise.

Expert, referent and increasingly informational power are owned. Nobody can take back what you know or how people regard you. These arrive with you at the new organisation.

The practical consequence is a straightforward investment argument. Positional power is worth using and not worth building, because building it means investing in something that will be repossessed. Personal power is slower to acquire and permanent.

This also explains why people without titles can be highly influential, and why some senior people are strangely ineffective the moment they need cooperation they cannot mandate. The person with no authority who everyone consults has expert and referent power. The executive whose requests are technically obeyed and quietly slow-walked has legitimate power and nothing else.

A worked example: the same request, six ways

A concrete case. You need another team to change their release schedule, which is inconvenient for them and important for you. You have no authority over them.

Legitimate: "This came from the steering group, so it needs to happen." Works if the claim is true and the authority is recognised. Produces compliance without commitment, and if the steering group's attention moves, so does their priority.

Reward: "If you can do this, I will make sure your team gets first call on the new infrastructure budget." Works while you can deliver. Creates a transaction, which means next time there will be a price.

Coercive: "If this slips, I will have to raise it with your director." Fast, and expensive. You have converted a colleague into someone managing a threat. The schedule may change. The relationship has, too.

Expert: "I have run this sequence three times. If we release in that order, the rollback path breaks, and here is what happened the last time." Works when your expertise is real and relevant, and it builds standing each time it proves correct.

Referent: "I would not ask if it were not genuinely needed." Works only if you have a history of not asking unless it is genuinely needed. It is drawing on years of prior behaviour, in one sentence.

Informational: "The client has moved their launch, which has not been announced yet, and it changes what the dependency looks like." Works by changing what they know, so the decision becomes theirs rather than yours.

Look at what separates them. The first three ask the other team to comply. The last three change what compliance means: expert power changes their understanding of the risk, informational power changes their picture of the situation, referent power draws on a belief that your judgment is trustworthy. In all three, the other team ends up wanting a different outcome rather than being made to accept one.

Why coercive power costs more than it returns

Coercion deserves separating out because it is the one base that actively destroys the others.

It produces compliance under observation and non-compliance the rest of the time. Once someone is managing a threat rather than pursuing an outcome, their attention shifts to appearing compliant, which is not the same as being effective and is frequently cheaper for them.

It also converts problems into secrets. People who expect punishment for bad news stop delivering bad news early, which is precisely when it is cheapest to fix. A great deal of organisational failure is the compounding of this one effect.

And it spends referent power to buy compliance, at a bad exchange rate. Every coercive act reduces the regard you are held in, which reduces the base that would have produced willing cooperation later.

There is a narrow legitimate use. Consequences must exist for genuine failures of conduct, and a system without any is not kind, it is unserious. The distinction is between consequences applied predictably and transparently for real breaches, which is fair process, and threat used as an instrument to get ordinary things done, which is what erodes everything.

How do you build referent and expert power deliberately?

Both are buildable. Neither is fast.

Expert power comes from depth plus the demonstration of it. Depth alone is invisible. The components are: go genuinely deep in a domain that matters to decisions rather than one that merely interests you; be right in public on things that are checkable, since credibility accrues when a prediction can be verified; say clearly when you do not know, because calibrated uncertainty is what separates expertise from confidence; and make the knowledge usable, since expertise that cannot be applied by the person receiving it produces admiration rather than influence.

Referent power comes from consistency over time, which is why it cannot be accelerated. The components are: do what you said you would do, reliably, including on small things where nobody is checking; represent people's interests when they are not present, which is the single most reliable generator of regard; take responsibility when something goes wrong rather than distributing it; and be useful without an immediate return, repeatedly.

The last one is worth dwelling on. Referent power is largely an accumulation of instances where you were helpful with no visible benefit to yourself. There is no shortcut, because the whole signal is that you did it without a reason, and doing it in order to accumulate referent power is a reason.

The problem with slow-building skills

Expert and referent power both take years, produce no visible feedback for most of that time, and are therefore exactly the kind of thing people intend to work on and never do.

The Leadership, Career & Influence world is structured around that gap. Daily quests, points and a streak keep five-minute lessons running through the stretch where nothing external is happening, with a streak shield for the day that goes wrong. Quizzes follow each topic and each course ends with a ten-question final exam, so you can see progress in the learning even while the real-world effects are still years out. Pass the final and you claim a verified certificate with your name on it, and certified learners join the expert network that answers other people's questions, which is itself a small piece of demonstrated expertise.

Diplomacy & influence runs fourteen courses alongside Networking at 36, and a Circle gives you a small group with a shared weekly goal and live sessions where everyone runs the same lesson together.

What is the difference between influence and manipulation here?

The framework describes mechanisms, and mechanisms can be used either way, so the question is fair and deserves a direct answer rather than a disclaimer.

The clearest test is what happens to the other person's information. Influence improves it: expert power gives them a better model of the risk, informational power gives them facts they lacked, referent power lets them borrow judgment they have good reason to trust. In each case they end up better equipped to decide.

Manipulation degrades it: fabricated urgency, selectively withheld context, implied authority that does not exist, manufactured obligation. The person ends up acting on a false picture.

A second test, useful because it is harder to rationalise around: would you be comfortable if they could see exactly what you did and why? Expert and referent power survive that inspection easily. Someone who learns you influenced them because you knew more and had a track record of being right is not upset. Someone who learns you invented a deadline is.

Informational power is the one to watch, because it sits closest to the line. Sharing information that changes someone's decision is legitimate. Controlling access to information so that only your preferred conclusion is reachable is not, even though both are technically informational power.

What to take from this

Influence has six sources, not one, and treating authority as the whole of it explains why people without titles are sometimes more effective than people with them.

Three are lent by your position and repossessed when it changes. Three are yours and travel. The investment case is obvious once stated and almost universally ignored.

Coercion is the fast option and it spends the other bases to buy compliance, converts problems into secrets, and compounds negatively.

Expert power is built by being deep, right in public, and honest about uncertainty. Referent power is built by consistency, by representing people when they are absent, and by being useful without an immediate return.

And the diagnostic is simple enough to run on yourself this week: when people do what you ask, do they do exactly that and stop, or do they extend it? The answer tells you which base you have been operating from.