You have seen the better idea lose. A sharper plan, more evidence, a cleaner business case, beaten by something vaguer that happened to have the right people behind it. That is not a failure of logic; it is a failure to read where the power sat and who needed bringing along. Organisational politics is simply what happens when smart people disagree about scarce resources, and the leaders who get things done are the ones who stop resenting that and start reading it.

The quick version

  • Power is the ability to get things done, especially against resistance. It comes from more than your job title, French & Raven mapped five everyday sources of it, only one of which is your formal position.
  • Politics is the activity of acquiring and using power to influence decisions when resources are scarce and people disagree. It is unavoidable, not inherently dirty, the ethics live in how you use it, not whether you engage.
  • Political skill is learnable and measurable. Researchers found it has four parts: reading people, influencing them, building networks, and coming across as genuine while doing it.
  • The honest caveat: politics tips into something corrosive when it serves you at the expense of the organisation. The test is whether your manoeuvring advances a shared goal or only your own standing.

The idea in depth

Most discomfort with "politics" comes from a definitional muddle. We use the word for two different things: the ordinary work of influence (legitimate, even admirable) and self-serving manipulation (genuinely toxic). Conflate them and you reach the worst possible conclusion, that the honest move is to opt out. Opting out doesn't make the politics stop; it just means it happens without you in the room. So separate the two, and get fluent in the first.

Where power actually comes from

The foundational map here is old and still unbeaten. In 1959, social psychologists John French and Bertram Raven set out five bases of social power: legitimate (authority from your position), reward (you control things people want), coercive (you can punish), expert (you know things others need), and referent (people identify with, respect or like you). Raven later added a sixth, informational power, control over a specific, persuasive piece of information. The point that matters for a working leader: only the first three travel with your job title, and they are the weakest in practice. Expert and referent power, credibility and relationships, are what move people who don't report to you, which is almost everyone whose help you actually need.

Audit your own power honestly, then, before your next big push. If your plan only works when people have to comply, you're leaning on legitimate and coercive power, the kind that produces grudging minimums and quiet resistance. The months before you need something are when you build the expert and referent power you can't requisition: be visibly useful, be reliable, be the person whose read on a problem others seek out. That account pays out exactly when formal authority runs short.

flowchart TD
  P(["Your power to get
something done"]) --> POS(["Positional
travels with the title"])
  P --> PER(["Personal
you have to earn it"])
  POS --> L(["Legitimate
authority of the role"])
  POS --> R(["Reward
you control what people want"])
  POS --> C(["Coercive
you can punish"])
  PER --> E(["Expert
you know what others need"])
  PER --> RF(["Referent
people respect / identify with you"])
  PER --> I(["Informational
a specific persuasive fact"])
French & Raven's bases of power, split into the kind your title gives you and the kind you have to build. The personal sources are what move people who don't report to you. Leaders Loop

Why some teams have more power than their headcount suggests

Power isn't only personal; it pools around certain roles and departments for structural reasons. The clearest account of this is the strategic contingencies theory set out by David Hickson and colleagues in "A Strategic Contingencies' Theory of Intraorganizational Power" (Administrative Science Quarterly, 1971). Their argument: a subunit gains power when it controls something the rest of the organisation critically depends on. Three things amplify that power, how much it copes with uncertainty the organisation is anxious about, how central its work is to everyone else's, and how non-substitutable it is (whether anyone else could do the job). It explains why, in one company, the legal team holds real sway and in another it's an afterthought: it depends on which uncertainty is currently keeping leadership awake.

Read the terrain structurally, then, not just personally. Before you ask why a particular function keeps winning the argument, ask what uncertainty it absorbs for the people above you. If you want your own team to carry more weight, the durable answer isn't louder advocacy. It's to attach your work to a problem the organisation is genuinely worried about, and to become the team that reliably copes with it. Influence follows criticality.

Politics as a normal part of the job, not a character flaw

The case for treating politics as legitimate work comes most directly from Stanford's Jeffrey Pfeffer, whose Managing With Power (1992) argues that knowing how to get things done, building support, overcoming resistance, assembling coalitions, is an essential and underrated component of leadership, not an embarrassing add-on. Pfeffer's blunter, more recent framing in 7 Rules of Power (2022) is that talent and being right are not self-executing; they need power behind them to land. He is describing the world as it is, which is the value of him: pretending influence doesn't decide outcomes leaves the influence to people less troubled by your scruples.

Opting out of the politics doesn't make it stop. It just means it happens without you in the room.

An honest limitation. This is where intellectual honesty matters, because Pfeffer's work is genuinely contested, critics argue that a power-first reading can read as licence for self-promotion and rule-breaking, and that it understates how much trust and competence still matter over time. Both can be true. Use the descriptive insight (influence is real, ignoring it is naïve) without swallowing the normative slide (therefore do whatever advances you). The line, drawn for leaders by Karen Dillon in the HBR Guide to Office Politics (2014), is to be constructively political, to read the dynamics and work them for mutual advantage and the good of the enterprise, rather than to grab. Politics that serves a shared goal is leadership; politics that only serves you is the thing everyone's afraid of.

The skill is learnable, and it has four parts

If this still sounds like a fixed trait, you've either got it or you haven't, the evidence says otherwise. Gerald Ferris and colleagues built and validated the Political Skill Inventory (Journal of Management, 2005), now the most widely used measure of the construct, and found political skill resolves into four learnable dimensions: social astuteness (reading people and situations accurately), interpersonal influence (a flexible, calibrated style that brings people along), networking ability (building useful relationships before you need them), and apparent sincerity (coming across as genuine and without hidden agenda). In their studies, political skill predicted managers' performance ratings, and, tellingly, it was unrelated to raw intelligence. This is not about being clever. It's about being attuned.

Each dimension is a muscle, not a verdict. Weak on social astuteness? Spend the first ten minutes of a contentious meeting saying nothing and mapping who defers to whom. Weak on networking? Build the relationship a quarter before you need the favour, when it costs nothing and reads as genuine. And notice the quiet warning inside "apparent sincerity": the dimension only works when the sincerity is real. People are good at detecting performed warmth, and a reputation for hidden agendas is the one political liability that doesn't recover.

A worked example

Take a product lead, call her Priya, who wants the company to consolidate three overlapping internal tools into one platform. (Illustrative scenario; the people and figures are a teaching example, not a real case.) The business case is strong: she models, say, an illustrative £400k a year in saved licence fees and engineering time. She presents it to the leadership team, expecting applause. It dies in the room.

What killed it wasn't the numbers. Reading it through the lens above: each of the three tools had an owner, and consolidation meant two of them losing their patch, a structural-power problem, not a logic problem. The finance director, whose uncertainty was next year's cost base, was the one person whose problem her plan actually solved, and Priya had never spoken to him before the meeting. She had legitimate power over her own team and none over anyone else's, and she'd spent her referent and expert capital on a slide deck instead of on people.

flowchart LR
  A(["Strong business case
~£400k saving (illustrative)"]) --> B{"Read the politics
before the meeting?"}
  B -->|"No, pitched cold to the room"| C(["Two tool owners lose
their patch, no allies
plan dies"])
  B -->|"Yes, mapped power first"| D(["Pre-wire finance as sponsor
give the losers a role
arrive with allies"])
  D --> E(["Decision goes through
because the room was
already won"])
The same case, two outcomes. The work that decides it happens before the meeting, not in it. Leaders Loop

The second attempt is a different piece of work. Priya maps the stakeholders by power and interest before she pitches anything (the stakeholder mapping discipline does exactly this). She takes the finance director to coffee and lets his cost problem frame the proposal, turning him into a sponsor with positional power she lacks. She gives the two losing tool owners a real role in the new platform's design, so consolidation reads as promotion rather than defeat. By the time the decision meeting happens, the room is already won; the meeting only ratifies it. Same numbers, opposite result, because the second time she did the political work the first version skipped.

Frequently asked questions

Isn't "playing politics" just a polite word for manipulation?

Sometimes, which is exactly why the distinction matters. Manipulation hides its intent and serves the manipulator at others' expense. Constructive politics is transparent about what you want and seeks an outcome others can also win from. The honest test, per Dillon's HBR Guide, is whether your influencing advances a shared goal or only your own standing. If you'd be comfortable explaining your move to the people affected by it, you're on the right side of the line.

I'm an introvert and I hate this. Am I doomed to lose?

No. Ferris's research found political skill is unrelated to general intelligence and, importantly, isn't the same as extroversion, its core is social astuteness, which is observation, not performance. Quiet, observant people are often better at reading a room than the loudest person in it. Networking can be one genuine relationship at a time rather than working a crowd. The skill rewards attention more than charisma.

How do I read the power in a new organisation quickly?

Watch decisions, not org charts. Notice who gets interrupted and who doesn't, whose objection ends a discussion, which meetings the real decisions actually happen in (often not the official one), and which team people check with before committing. Strategic-contingencies theory gives you a shortcut: ask what the organisation is currently anxious about, and whoever absorbs that uncertainty holds more power than their title implies.

What if the politics where I work really is toxic?

Some environments are genuinely corrosive, and naming that honestly is its own form of clarity. Constructive engagement has limits; where the unwritten rules reward dishonesty and punish candour, the skill becomes protecting your team, documenting decisions, and deciding how long you're willing to stay. Reading politics well includes reading when a place can't be navigated, only endured or left.

Where does power come from if I don't have a senior title?

From the sources that don't require one. French & Raven's expert and referent power, being demonstrably good at something the organisation needs, and being someone people trust, are available at any level, and Hickson's work adds a structural route: attach yourself to the uncertainty leadership most wants resolved. Title is the least durable form of power; the others you can start building today.

Related in the Toolkit

Reading the room is the diagnosis; the moves that follow have their own modules. Once you've mapped who holds power (stakeholder mapping & analysis), the next job is usually building the coalition that carries a decision, and doing all of it without losing the self-awareness that keeps you on the constructive side of the line.

Where to go next