The Psychology of Power: What Happens to the Brain When You Gain Authority.

In 1963, Stanley Milgram brought ordinary people into a laboratory at Yale University and asked them to deliver what they believed were potentially lethal electric shocks to a stranger.^[1] Sixty-five per cent complied.^[2] The finding, one of the most replicated in the history of psychology, was not about cruelty. It was about what happens to the human brain when someone else holds the authority. The person pressing the button did not become evil. They became obedient. And obedience to authority, Milgram understood, was not a character flaw. It was a neurological event: the presence of an authority figure restructured how participants processed moral information, how they weighted consequences, how they perceived the suffering of another person sitting behind a wall. Kipnis demonstrated the other side of this equation a decade later. Managers given arbitrary power over workers began to devalue those workers' contributions and increase psychological distance, the first empirical evidence that power corrupts the power-holder, not just the subordinate.^[3] The psychology of power does not begin with the powerful. It begins with the fact that human brains are built to recalibrate in the presence of rank.

That recalibration runs in both directions. The person who gains authority undergoes a transformation at least as profound as the person who defers to it. Over the past two decades, a convergence of neuroimaging, transcranial magnetic stimulation, neuroendocrine assays, and large-scale behavioural experiments has documented what that transformation looks like at the level of neural circuits. Power activates the brain's behavioural approach system: the motivational architecture that drives reward-seeking, risk tolerance, and goal pursuit.^[4] It suppresses the motor resonance system that allows one brain to simulate what another brain is feeling.^[12] It weakens the coupling between the brain regions that compute self-interested valuation and the regions that apply moral brakes.^[17] The result is not a person who has chosen to stop caring. It is a person whose neural hardware has quietly made caring harder.

The scale of the evidence is no longer debatable. A 2025 meta-analysis published in Psychological Bulletin synthesised 1,712 effect sizes drawn from 813 independent samples, 432 manuscripts, and 269,534 participants.^[21] The conclusion: the approach-inhibition theory of power (the framework proposing that elevated power activates approach and suppresses inhibition) is broadly confirmed across two decades of empirical testing. Power changes how the brain operates. The question that remains is what, precisely, to do about it.

The reason this matters for anyone who leads (or works under someone who does) is that the transformation is automatic. Hogeveen, Inzlicht, and Obhi's landmark TMS study demonstrated that the mirroring suppression was not a deliberate choice or a side effect of social distance.^[12] It was a default neurological response to holding power. The motor cortex simply stopped simulating others' actions with the same fidelity. This is not a metaphor. It is a measurement. And it means that the leader who insists they still "get" their team may be reporting an experience their motor cortex no longer supports.

Dacher Keltner, whose research group at Berkeley has studied the psychology of power for over two decades, calls this the power paradox: the social skills that win you influence (empathy, attunement, generosity) are the first casualties of having it.^[22] The paradox is not philosophical. It is neurochemical. Power triggers dopamine and testosterone surges that orient the brain toward reward and away from threat.^[9][23] It reduces the cortisol vigilance that keeps social behaviour cautious and calibrated. The hormonal climate shifts, and the cognitive architecture follows.

What makes this especially dangerous is that the person undergoing the change is usually the last to notice. Power reduces metacognitive accuracy: the ability to assess one's own cognitive performance.^[6] The powerful believe they are reading the room when they are reading themselves. They believe they are taking perspectives when they are projecting. The transformation is invisible from the inside, which is precisely why it requires structural countermeasures rather than good intentions.

The mechanism begins with a motivational shift. When a person gains power (through promotion, election, or social dynamics) the brain's behavioural approach system (BAS) activates.^[4] This is not a subtle psychological preference. It is a measurable change in neural activity. The BAS drives reward-seeking, risk tolerance, goal pursuit, and action orientation. It is mediated by dopaminergic pathways that connect the ventral striatum, prefrontal cortex, and amygdala, the same circuits that process monetary reward, sexual attraction, and substance-related pleasure.^[18][13][33] Shared neural coding for social hierarchy and reward value has been documented in primate amygdala, confirming that rank and reward are processed through overlapping circuits.^[33] Keltner, Gruenfeld, and Anderson formalised this in 2003 as the approach-inhibition theory: power activates approach; powerlessness activates inhibition.^[4] Sixteen years of subsequent testing, culminating in the 2025 meta-analysis of 1,712 effect sizes, confirms the framework is robust.^[21]

The practical consequence is immediate. Power-holders become more action-oriented: more likely to take a card in a risky game, to act against an environmental nuisance, to choose quickly in a social dilemma.^[5] Guinote's 2017 review showed that power energises thought, speech, and action; it orients cognition toward salient goals and magnifies self-expression while reducing attention to peripheral information.^[15] That matters because the "peripheral information" power filters out includes other people's emotional states, knowledge gaps, and perspectives. The cognitive efficiency that power delivers comes at the cost of social perception.

The neurochemistry is equally telling. Mehta and Josephs demonstrated that testosterone's relationship to dominant behaviour depends on cortisol: testosterone predicts dominance only when cortisol is low.^[9] High cortisol (the signature of chronic stress) blocks the hormonal conditions for effective authority. In non-human primates, Sapolsky's 25-year Amboseli baboon dataset showed that high-ranking animals with stable hierarchies had both the lowest cortisol and the lowest atherosclerosis.^[18] Human research on rank and cortisol suggests a parallel: the physiology of effective power requires a specific dual-hormone climate, the interaction between the hypothalamic-pituitary-gonadal axis (HPG, governing testosterone) and the hypothalamic-pituitary-adrenal axis (HPA, governing cortisol), not just raw dominance drive.^[30]

The second stage of the loop is where the damage becomes personal. As the approach system fires, the brain's mirror neuron system (the motor cortex circuitry that simulates observed actions) goes quiet. Hogeveen, Inzlicht, and Obhi's 2014 experiment is the cleanest demonstration.^[12] Using transcranial magnetic stimulation, they measured motor-evoked potentials (MEPs) while participants watched a hand squeezing a ball. Participants primed into high-power states showed significantly lower MEP amplitudes than those in low-power states. The motor cortex was no longer resonating with what it saw. The mirroring suppression was automatic, not mediated by social distance or explicit motivation. It was a default neural consequence of power.

This matters because motor resonance is a foundation of empathic accuracy. When your motor cortex simulates another person's actions, it contributes to your ability to infer their intentions, read their emotions, and predict their behaviour.^[12] When that simulation quiets, other people become less legible. Galinsky and colleagues showed the downstream effect across four experiments: high-power participants drew the letter "E" in a self-oriented direction (failing to consider the viewer's perspective), anchored excessively on their own knowledge states, and showed significantly impaired emotion recognition.^[6]

The combined effect is a leader who is neurologically less equipped to understand the people they lead and who simultaneously becomes more confident in their social judgments. Power increases what psychologists call illusory transparency (the belief that one's internal states are obvious to others) while simultaneously reducing the accuracy of social perception.^[6][15]