Every generation rediscovers the same warning. Lord Acton put it in a letter in 1887: power tends to corrupt. Office gossip says the same thing, less elegantly, about the colleague who got promoted and stopped returning calls. For most of human history, that was as far as the explanation went — a moral verdict, delivered after the fact, about someone’s character.

Psychologists have spent the last twenty-five years testing whether the folk wisdom holds up. What they have found is stranger, and more specific, than simple corruption.

ONE IDEA

Power does not usually reveal a person’s true character. It changes what their mind does — measurably, predictably, and in most people, not just the ones we’d already suspect.

The clearest account of this comes from the psychologist Dacher Keltner at the University of California, Berkeley, who spent decades studying how people gain and lose influence. His finding, sometimes called the power paradox, is this: the qualities that get a person into a position of power — reading a room well, empathizing with others, generosity, restraint — are often the first things that position erodes once they hold it.

There is a mechanism behind this, not just an observation. Keltner and his colleagues Deborah Gruenfeld and Cameron Anderson proposed what they called the approach/inhibition theory of power: having power activates a psychological “approach” system associated with reward-seeking, risk-taking, and disinhibited action, while lacking power activates an “inhibition” system tuned to threat, social judgment, and caution. In plain terms, power presses the gas pedal on action and eases off the brake that comes from worrying what other people think. That shift shows up in how freely powerful people speak, how quickly they act, and how little they check themselves before doing either.

A related line of research gets at exactly what gets lost. In a well-known 2006 study, the psychologist Adam Galinsky and colleagues found that people who were primed to feel powerful were worse at seeing a situation from someone else’s point of view — literally, in one experiment, more likely to draw a letter “E” on their own forehead facing themselves rather than oriented so an observer could read it, a small but telling sign of self-referential thinking crowding out the instinct to imagine another person’s vantage point.

What people commonly get wrong is treating this as proof that power “reveals” someone’s hidden true self — that the corrupted boss was secretly like that all along. The research suggests something less comfortable: power changes cognition and behavior in most people who hold it long enough, largely regardless of the decent intentions they started with. That reframes the old question. It is not mainly “how do we screen out the bad apples before we hand them power,” though that still matters. It is “how do we build structures that counteract what power predictably does to anyone,” because anyone, eventually, is the honest answer.

Evidence strength: strong. The approach/inhibition theory of power has held up and been extended across two decades of behavioral research. A few individual findings inside this broader literature — including some specific versions of the perspective-taking effect — have had mixed results when other labs tried to replicate them, which is exactly why this should be read as a strong, well-supported pattern across many people rather than an iron law that predicts any one individual with certainty.

ONE METHOD

If power reliably narrows attention toward yourself and away from other people, the countermeasure has to work against that pull deliberately, not rely on remembering to be humble. A simple three-step check does this:

1. Name the pull. Notice when you feel unusually certain, unusually fast, or unusually unwilling to hear an objection. That feeling is not proof you’re right. It is often just the approach system doing what it does under power.

2. Borrow a view. Before finalizing a decision that affects someone with less power than you, say or write one sentence describing how the situation looks from their side — not how you’d explain your reasoning to them, but what they are actually experiencing.

3. Install a check. Give one specific person explicit permission, in advance, to disagree with you — and treat that disagreement, when it comes, as information rather than a challenge to be managed.

Evidence strength: moderate. The core ingredient — deliberately taking another person’s perspective before deciding — has experimental support as a counterweight to power’s self-focusing effect; Galinsky’s own follow-up research found that prompting people to consider others’ viewpoints reduces exactly the bias described above. The three-step packaging here is a practical synthesis of that finding rather than a single tested intervention, so treat it as a sound method to try, not a guaranteed fix.

ONE ACTION

For the next seven days, before every decision where you hold more authority than the other person involved — approving or rejecting something, evaluating someone’s work, choosing on someone else’s behalf — pause and write one sentence describing the situation from their point of view before you decide. Keep a simple tally of how many decisions this changed, even slightly, by the end of the week. The action is not “be more empathetic.” It is this specific, repeatable habit, done a measurable number of times, this week.

ONE SHIFT

Before: decisions happen quickly, objections rarely reach you, and it feels like everyone mostly agrees. After: you start to notice the half-second pause before you would have acted on instinct, and — if the check is working — the people around you start disagreeing with you more often, out loud, before a decision is finalized rather than after it goes wrong. That second change is the one worth watching for. An increase in visible disagreement is usually a sign the system is healthier, not that something has gone wrong.

ONE QUESTION

When was the last time someone with less power than you told you “no,” or disagreed with you to your face? If the honest answer is that you can’t remember, ask what that silence actually means. It is rarely proof that everyone genuinely agrees with you. More often it is the first visible sign that power has already begun narrowing what you’re able to hear.

ONE SENTENCE

Power rarely corrupts character on its own — it quietly redirects attention toward yourself and away from everyone else, unless you deliberately fight to keep looking outward.

ONE SCENARIO

Consider a software engineer promoted to team lead after years as the most collaborative person on her team — the one who always asked quieter colleagues what they thought before a decision shipped. Six weeks into the new role, she starts making calls alone. She skips the open discussion she used to run before every release. When a junior engineer flags a risk in an upcoming deployment, she thanks him for the note and ships anyway. The deployment fails in production, and the postmortem finds the junior engineer’s warning was accurate and specific.

What should she do, and why? The evidence-informed answer is not “try harder to be humble” — approach/inhibition theory suggests that under sustained authority, good intentions alone are a weak defense against a fairly predictable cognitive shift. The stronger response is structural: reinstate the practice that made her effective before the promotion, but make it a formal role rather than a personal habit — assign someone, on a rotating basis, whose explicit job in every release review is to raise the strongest objection before a decision ships, and treat a lack of any objection as a signal to look harder, not a green light. That rebuilds, deliberately, the perspective-taking step that power was already working to erode — not because she became a worse engineer or a worse person, but because the research suggests almost anyone in her position would have started drifting the same way.

Explore more from Specialty Digest’s Big Ideas series, or subscribe to Specialty Digest Weekly for the ideas shaping how we think.