Does Power Corrupt? Psychology, Morality, Behavior, and the Power Paradox
Updated: 3 days ago
Author: Ukrainian Psychological Hub · Published: September 24, 2026 · Editorial Policy
Power can corrupt behavior, but psychology does not support the slogan as a deterministic law. The strongest modern reading is more conditional: power changes the psychological environment in which people act. It reduces dependence on others, increases freedom of action, strengthens approach-oriented motivation, and can make active goals, motives, identities, and impulses easier to express. Those changes can produce selfishness, moral blind spots, objectification, overconfidence, or abuse in some people and situations. They can also support responsibility, principled action, helping, and difficult prosocial decisions in others.
A 2025 meta-analysis of the approach–inhibition theory synthesized 1,712 effect sizes from 813 independent samples across 432 manuscripts and 269,534 participants. It concluded that the theory is broadly supported: power is associated more strongly with approach-oriented outcomes such as attention to rewards, positive affect, and disinhibited behavior than with inhibition-oriented outcomes, and the behavioral effects appear to operate partly through changes in attention, cognition, and affect (Sabey et al., 2025).
That finding helps explain why the question “Does power corrupt?” has resisted a simple yes-or-no answer. Power does not contain one moral direction. It changes constraints, attention, confidence, goal pursuit, and social distance. What follows depends on the person, the goal, the role, the norms, the legitimacy of the position, the surrounding status structure, and the accountability system (Guinote, 2017; Foulk, Chighizola, & Chen, 2020).
Dacher Keltner’s “power paradox” gives this tension a memorable public formulation: people can gain influence through qualities that benefit others, yet the experience of power can weaken some of the attention, restraint, and social sensitivity that helped them gain influence in the first place. The power paradox is best treated as an integrative psychological interpretation of a research program, not as a universal experimental law (Keltner, 2016).
Does Power Corrupt? The Short Psychological Answer
Sometimes, under identifiable conditions. Power is associated with psychological processes that can facilitate corrupt or antisocial behavior: greater approach motivation, less inhibition, more action, greater confidence, reduced dependence on other people, greater social distance, reduced spontaneous perspective taking, and a stronger tendency to pursue focal goals. If the goals and norms active in the situation are self-serving, exploitative, or unethical, these mechanisms can make harmful behavior more likely (Keltner, Gruenfeld, & Anderson, 2003; Guinote, 2017; Sabey et al., 2025).
Yet power can also facilitate ethical or prosocial behavior. Research shows that a strong moral identity can reverse the usual self-interest pattern, that responsibility-focused interpretations of power improve attention to others, and that experimentally or naturally elevated power can in some contexts support helping that brings little or no status benefit (DeCelles et al., 2012; Scholl, 2020; Gobel, Choi, & Uchida, 2025; Hadar et al., 2026).
So the most defensible conclusion is not that power always corrupts, nor that power merely reveals a fixed personality. Power changes the conditions under which motives become behavior. It often acts as an amplifier, a disinhibitor, and a source of psychological freedom. Whether that freedom becomes exploitation or responsibility is a question about both the person and the system.
What Psychologists Mean by Power
In social psychology, social power is usually defined in relational terms: one actor has greater power when they can shape another person’s valued outcomes, often because they control resources, decisions, sanctions, opportunities, or alternatives. A fuller treatment belongs to our article Social Power: Definition, Types, French and Raven, Keltner, and Modern Psychology.
This definition matters because power is not the same thing as status. Status concerns respect, esteem, prestige, or standing in the eyes of others. A person can be powerful without being respected, or respected without controlling important outcomes. The distinction is central enough that studies sometimes find opposite psychological effects for power and status. See Power vs Status: Social Power, Prestige, Dominance, and Hierarchy.
Power is also not identical to authority. Authority is a recognized or legitimate right to direct, decide, or command within a role or institution. Power can exist without that legitimacy, and authority can be weak in practice when the role-holder lacks effective influence. See Authority in Psychology: Legitimacy, Expertise, Trust, and Obedience and Power vs Authority: Weber, Legitimacy, Influence, and Obedience.
Finally, power over other people is not the same thing as personal control over one’s own life. That distinction is especially important when interpreting claims about whether high positions are psychologically harmful or beneficial.
What Does “Corrupt” Mean in Psychological Research?
“Corruption” is not a clinical diagnosis and does not describe a single psychological construct. In research on power, the word is used broadly for shifts toward self-serving, exploitative, unethical, aggressive, hypocritical, objectifying, or norm-violating behavior. Different studies operationalize these outcomes differently: resource allocation, cheating, moral judgment, exploitation, aggression, advice taking, interpersonal attention, or decisions made under moral hazard.
That breadth is one reason sweeping claims should be treated carefully. A study showing reduced perspective taking does not by itself show criminal corruption. A study showing self-interested allocation does not prove a permanent personality transformation. A study showing greater confidence does not prove narcissism. The scientifically useful question is narrower: which psychological processes change with power, and under what conditions do those changes increase harmful or prosocial behavior?
The Strongest Current Framework: Power Increases Approach and Reduces Constraint
The approach–inhibition theory of power proposed that elevated power is associated with rewards, freedom, and reduced social constraint, which activate approach-related tendencies; low power is associated more strongly with threat, punishment, dependence, and inhibition. The original framework predicted greater positive affect, attention to rewards, automatic cognition, and disinhibited behavior under higher power (Keltner, Gruenfeld, & Anderson, 2003).
For years, this influential theory accumulated both supporting findings and inconsistencies. The 2025 Psychological Bulletin meta-analysis is therefore unusually important. Rather than treating classic individual studies as the final word, it tested the broader theory across hundreds of manuscripts and found overall support, with approach associations larger than inhibition associations and affect playing a substantial mediating role (Sabey et al., 2025).
This is not a moral theory. Approach can serve selfish goals or collective goals. Disinhibition can release aggression or courage. Reduced concern about social consequences can enable exploitation or principled dissent. The mechanism predicts greater movement toward salient goals and fewer behavioral brakes; it does not specify that the destination must be bad.
Power Makes Action Easier
One of the most replicated themes in the literature is that power facilitates action. In classic experiments, participants assigned or primed to feel more powerful acted more readily in ambiguous situations, including situations where the action could have prosocial or antisocial consequences (Galinsky, Gruenfeld, & Magee, 2003).
Guinote’s major review framed power as an intensifier of goal-related approach motivation. Power holders tend to prioritize active goals, express active parts of the self more strongly, and organize attention and effort around what seems important in the moment. This can improve decisiveness and goal pursuit, but it can also narrow attention around personally salient aims when other people’s needs are not part of those aims (Guinote, 2017).
This helps explain why “power corrupts” can look true in some settings. A person with an exploitative goal may become more effective at pursuing it. A person with a protective or public-serving goal may also become more effective at pursuing that. Power changes the translation from motive into action.
Power, Perspective Taking, and Attention to Other People
A second major line of evidence concerns social attention. In a classic series of studies, higher power was associated with less spontaneous perspective taking and, in one experiment, lower accuracy when judging other people’s emotional expressions (Galinsky et al., 2006).
Related work found that people with a higher sense of power reported less distress and compassion when listening face-to-face to another person describe suffering, an effect the authors linked to differences in affiliation motivation rather than to a general inability to decode emotion (van Kleef et al., 2008).
Modern reviews complicate the simple claim that power destroys empathy. Power tends to reduce social attunement and attention in some contexts, but interpersonal accuracy depends heavily on how power is construed. When power is understood as responsibility rather than opportunity, accuracy can improve (Schmid Mast, Khademi, & Palese, 2020).
Recent preregistered research strengthens that contextual interpretation. Across three studies, people who construed their power as entrusted by others rather than achieved solely through personal merit showed greater other-orientation and more altruistic behavioral tendencies (Gobel, Choi, & Uchida, 2025).
The evidence therefore supports a risk, not an inevitability: power can weaken spontaneous attention to other minds, especially when the powerful person is relatively self-focused and less dependent on others. It does not mechanically eliminate empathy.
For a dedicated synthesis of perspective taking, compassion, emotion recognition, empathic accuracy, and social distance, see Power and Empathy: Perspective Taking, Social Distance, and Emotional Accuracy.
Social Distance: Why Power Can Make Other People Feel Less Psychologically Close
Power relations often create asymmetric dependence. The less powerful person typically has more reason to monitor the more powerful person, because the powerful person controls outcomes that matter to them. The more powerful person has less instrumental need to monitor downward. Social-distance theory proposes that this asymmetry contributes to greater psychological distance among the powerful.
A 2020 review concluded that higher power tends to reduce attunement and attention to others, increase objectification and dehumanization, and increase expressions of dominance or aggression under some conditions. It also emphasized that these effects are moderated by factors such as expectations, norms, and individual differences (Magee, 2020).
This asymmetry is psychologically important because many moral failures begin before an explicit moral judgment is made. If another person’s needs, constraints, emotions, or individuality receive less attention, harmful consequences can become less salient during decision-making. For the narrower evidence on category-based judgment, attention, and individuation under power, see Power and Stereotyping: Attention, Individuation, and Social Judgment.
Power and Objectification: Seeing People as Means to Goals
Power can also change how people represent other people. Across six studies, Gruenfeld and colleagues found that power increased instrumental attention to social targets: powerful participants’ approach toward others depended more strongly on how useful those people were for active goals (Gruenfeld et al., 2008).
Objectification here does not mean that every powerful person consciously dehumanizes subordinates. The mechanism is subtler. When goal pursuit becomes dominant and another person is relevant mainly as a route to that goal, individuating information can become less important. In organizations or intimate relationships, repeated instrumental treatment can become one pathway through which asymmetrical power produces disrespect or exploitation.
This is also why Power Dynamics: Psychology of Influence, Dependence, Status, and Control is a separate concept from this article. Power dynamics describe the structure of mutual influence and dependence; “corruption” describes possible behavioral consequences of holding or experiencing power.
Power, Self-Interest, and Moral Identity
One of the clearest challenges to the idea that power has a single moral effect comes from research on moral identity. DeCelles and colleagues found that power was associated with more self-interested behavior among people with weaker moral identity but less self-interested behavior among people with stronger moral identity. Their proposed mechanism was moral awareness: power appeared to decrease moral awareness for weak-moral-identity participants and increase it for strong-moral-identity participants (DeCelles et al., 2012).
This result supports an amplification account. Power can make it easier for existing motives and identities to organize behavior. A person whose self-concept strongly includes being moral may use greater freedom to act consistently with that identity. A person whose active priorities center on personal gain may use the same freedom differently.
It also shows why the question “Does power reveal the real person?” is too simple. Power does not merely reveal a fixed inner essence. It changes attention, felt constraint, confidence, and action tendencies while interacting with preexisting identity and context. The result is person-by-situation interaction, not pure revelation.
Power and Moral Reasoning
Some evidence indicates that power can shape not only behavior but the reasoning that precedes it. In an incentivized experiment, Giurge and colleagues separated two components of leader power: power over more versus fewer followers and power to enforce one’s will through greater discretion over outcomes. Power over more followers reduced principled moral reasoning, while greater discretionary power enabled more self-interested behavior (Giurge et al., 2021).
Yet the broader literature does not support the claim that power simply shuts down morality. A review of power and moral thinking argued that power can amplify multiple forms of moral thought, including deliberation, rule-based thinking, and integration of feelings and cognition (Fleischmann & Lammers, 2020).
These findings can coexist. Power can increase the intensity or autonomy of moral cognition while still biasing which concerns are salient, whose outcomes receive attention, or how self-interest is justified. Moral judgment, moral attention, and moral behavior are related but distinct outcomes. The narrower evidence on disinhibition, self-interest, moral identity, responsibility, and ethical conduct is developed in Power and Morality: Disinhibition, Self-Interest, Responsibility, and Ethics.
Moral Hypocrisy and Legitimacy
An influential set of experiments found that participants in high-power conditions could become more demanding when judging others’ transgressions while being more permissive toward their own, a pattern described as moral hypocrisy. Importantly, the same research found that legitimacy changed the effect: illegitimate power reversed the pattern in the final study (Lammers, Stapel, & Galinsky, 2010).
The article was later explicitly listed by Psychological Science among Stapel-coauthored papers that the Levelt investigation found no evidence of fraudulent data or practices in. Even so, the scientific lesson is broader than one famous study: legitimacy, expectations, and social norms can alter how power is experienced and expressed.
Other experiments similarly show that prescriptive expectations matter. When participants were reminded how powerful people should behave, power could reduce cheating relative to low power; when descriptive expectations that powerful people behave badly were made salient, the pattern went the other way (Hu, Rucker, & Galinsky, 2016).
Power, Confidence, Overconfidence, and Risk
Power often increases confidence and action readiness. Those effects can be useful when hesitation is costly, but confidence can exceed accuracy. Across five experiments, Fast and colleagues found that experiencing power increased overconfident decision-making and could produce monetary losses. The effect depended on subjective sense of power: objective power predicted overconfidence when it translated into feeling powerful (Fast et al., 2012).
Overconfidence is not corruption in itself. It becomes relevant because power holders frequently make decisions that affect other people. When confidence rises faster than accuracy, consultation can fall, warning signals can be discounted, and the cost of error can spread downward through a hierarchy.
This is one reason the broad psychology of power cannot be reduced to morality. Some harms arise through self-interest; others arise through attention failures, excessive certainty, insufficient feedback, or strong goal pursuit combined with weak correction.
Power Over Others Is Not the Same as Personal Control
People often seek higher positions partly because they want autonomy and control over their own lives. Psychological research suggests this motive should be separated from control over other people.
Cislak and colleagues found that power over others and personal control had different interpersonal correlates. Across studies, power over others positively predicted aggressiveness and exploitativeness, whereas personal control predicted these outcomes negatively (Cislak et al., 2018).
This distinction prevents a common interpretive error. Greater autonomy, competence, and self-determination are not equivalent to having leverage over other people. A person can gain personal control without gaining coercive or resource-based power, and the psychological consequences may differ substantially.
Does Power Attract Corruptible People, or Does It Change People?
Both selection and transformation are plausible, and real-world hierarchies contain both. Certain personalities and motives can affect who seeks, obtains, or retains power. At the same time, experiments that randomly assign or manipulate power show that the experience of power itself can alter cognition, affect, and behavior. That makes a pure selection explanation insufficient.
The experimental literature is especially valuable here because it demonstrates situational effects independent of who initially sought power. The evidence on action orientation, perspective taking, objectification, moral awareness, and accountability includes manipulations of power rather than only comparisons between naturally powerful and powerless people.
However, laboratory power is not identical to years of institutional authority. Many studies use temporary roles, recall tasks, resource asymmetries, or subjective sense-of-power manipulations. These methods provide causal leverage but do not reproduce the full ecology of real organizations, families, professions, or political institutions. Reviews therefore increasingly emphasize context and the way power is construed rather than assuming one uniform effect (Foulk, Chighizola, & Chen, 2020).
What Is the Power Paradox?
The “power paradox” is Dacher Keltner’s name for a recurring tension in social influence. People often gain influence by advancing group interests and displaying socially valued qualities such as generosity, fairness, empathy, cooperation, or emotional intelligence. Once power is acquired, however, reduced dependence and increased freedom can weaken some of the very social sensitivities that helped create influence (Keltner, 2016).
The idea is compelling because it joins two different moments in a power trajectory: acquisition and possession. The traits and behaviors that help a person become influential are not necessarily the same psychological conditions that prevail once that person can act with fewer constraints.
The paradox should not be converted into a mechanical sequence in which every prosocial leader inevitably becomes abusive. Its value is explanatory: it identifies a feedback problem. When power reduces the need to attend to others, the person may lose access to corrective social information precisely when their decisions affect more people.
That feedback problem also helps explain why power can become self-reinforcing. Subordinates may hesitate to disagree, powerful people may receive selectively positive information, and successful decisions can strengthen confidence even when success partly reflects structural advantages. The larger the asymmetry, the more deliberately feedback systems may need to be designed.
Status Can Change the Effect of Power
Power and status often travel together, but they are psychologically distinct. This matters for corruption. In three studies, Gu and colleagues found that power increased corrupt intent or behavior when status was low, whereas the effect disappeared when status was high (Gu et al., 2020).
One interpretation is that low-status power creates a particularly unstable combination: a person can control outcomes while lacking respect or recognition. The resulting experience may intensify efforts to exploit power or defend rank. The finding should not be universalized beyond the studied samples, but it reinforces a central lesson of the field: the effect of power depends on what other dimensions of hierarchy accompany it.
Recent work also finds that power and status can predict prosocial behavior differently. In studies published in 2026, elevated power was associated with more helping or harm-prevention behavior than elevated status when the prosocial act offered no status benefit and sometimes carried reputational cost (Hadar et al., 2026).
Responsibility Can Change What Power Does
Power can be construed as opportunity—freedom to make things happen—or as responsibility—an obligation to take care of people, tasks, or outcomes. A contemporary review argues that this distinction helps reconcile seemingly contradictory findings: powerful people may pursue their own opportunities, but they can also become highly task-focused or other-focused when responsibility is salient (Scholl, 2020). For the dedicated evidence on felt responsibility, leadership obligation, advice taking, and the psychological burden of consequential control, see Power and Responsibility: Leadership, Moral Obligation, and the Burden of Control.
The recent work on entrusted power points in the same direction. When people understand power as something conferred by others rather than simply earned by personal merit, they show more perspective taking and altruistic tendencies (Gobel, Choi, & Uchida, 2025).
This suggests a practical distinction between possessing power and interpreting what power means. Roles, institutional norms, selection systems, ceremonies, feedback practices, and leadership training can all influence whether a power holder thinks “I can do what I want” or “I am responsible for consequences that others cannot control.”
Accountability: One of the Most Important Boundary Conditions
Power becomes especially risky when decisions carry consequences for others but the decision-maker faces weak scrutiny, weak justification requirements, or no meaningful correction. Experimental research supports the importance of accountability, although accountability itself is not a single mechanism.
In three studies of decisions under moral hazard, Pitesa and Thau found that power increased self-serving decisions when appropriate accountability mechanisms were absent. Procedural accountability—requiring decision-makers to justify how they made the decision—reduced self-serving choices and curbed the negative effect of power (Pitesa & Thau, 2013).
The broader lesson is not that any oversight system automatically prevents abuse. Accountability works through specific expectations, audiences, procedures, and consequences. Poorly designed accountability can become ritualistic or encourage impression management. Effective safeguards make reasons inspectable, preserve independent feedback, and reduce the ability of one actor to suppress correction.
This institutional layer is crucial because a psychology of corruption that focuses only on the character of leaders misses the structure around them. Power is relational. The surrounding system determines how much dependence one person can create, how easily others can refuse, and whether errors or abuses can be challenged.
Does Absolute Power Corrupt Absolutely?
As a literal scientific proposition, no. Psychology does not establish that every person given sufficiently great power will inevitably become corrupt. “Absolute” is not a standard experimental variable, and corruption is not a unitary outcome.
The phrase remains psychologically insightful because extreme concentration of power can combine several known risk factors: reduced dependence on others, greater freedom of action, fewer external constraints, larger consequences of personal goals, weaker corrective feedback, and greater opportunity to shape the social environment itself.
The strongest evidence supports a probabilistic interpretation. As power increases and meaningful constraints weaken, the consequences of the power holder’s active goals and dispositions can become larger. That makes moral identity, responsibility, institutional design, accountability, independent information, and countervailing power more consequential—not less.
When Power Is Most Likely to Produce Harm
Across the literature, several conditions repeatedly appear as risk multipliers. They should be understood as patterns rather than as a diagnostic checklist.
• Weak accountability: decisions are difficult to review, challenge, or reverse.
• Strong self-interest or exploitative goals: power increases the ability to act on motives that are already active.
• High social distance: the people affected by decisions are psychologically or structurally remote from the decision-maker.
• Low dependence on feedback: the power holder can ignore dissent without meaningful cost.
• Instrumental views of others: people are treated primarily as means to goals rather than as sources of independent information and welfare.
• Overconfidence: subjective certainty grows faster than decision accuracy.
• Low status combined with high power in some contexts: research suggests this combination can increase corrupt behavior relative to high-power/high-status conditions.
• Opportunity-focused construals of power: power is experienced primarily as permission and freedom rather than as responsibility.
None of these conditions guarantees abuse. Together they help explain why the same nominal amount of power can have very different effects across people and institutions.
When Power Can Support Better Behavior
Power can also remove obstacles to prosocial action. The ability to allocate resources, protect others, enforce fair rules, or act despite social pressure can be morally useful. Several lines of research identify conditions under which this side of power becomes more visible.
• Strong moral identity can make power less self-interested rather than more self-interested (DeCelles et al., 2012).
• Responsibility-focused power can improve attention to others and task stewardship (Scholl, 2020).
• Entrusted-power construals can increase other-orientation and altruistic tendencies (Gobel, Choi, & Uchida, 2025).
• Prescriptive expectations about how the powerful should behave can reduce unethical behavior in experimental settings (Hu, Rucker, & Galinsky, 2016).
• Recent research separating status from power finds contexts in which elevated power predicts more costly prosocial behavior than elevated status (Hadar et al., 2026).
These findings do not turn power into a virtue. They show that power is a capacity whose moral direction is shaped by motives, identities, role meanings, and constraints.
Does Power Change Personality?
Power clearly changes psychological states and behavior. Evidence supports effects on attention, affect, confidence, action, social distance, and goal pursuit. Whether it permanently changes broad personality traits is a different question.
Many classic studies measure temporary experiences of power rather than long-term personality development. A powerful role held for years could plausibly reinforce habits, social networks, beliefs, and expectations, but the experimental literature is much stronger for immediate psychological processes than for irreversible personality transformation.
It is therefore more accurate to say that power can change how personality and motives are expressed. Guinote’s “active self” account is especially useful here: power tends to increase expression of the goals and self-aspects that are currently active (Guinote, 2017).
Does Power Cause Narcissism?
Power and narcissism are distinct constructs. Narcissistic traits can influence leadership emergence, power seeking, interpersonal behavior, and how power is used, but possessing power is not equivalent to developing narcissistic personality disorder or even to becoming high in trait narcissism.
Some behaviors associated with power—confidence, self-focus, reduced advice taking, entitlement, or social distance—can superficially resemble narcissistic behavior. That overlap does not justify treating power as a cause of a clinical disorder. The safer scientific claim is that power can create behavioral conditions in which self-focus and overconfidence become more visible or consequential.
What the Evidence Says About Empathy
The popular claim that “power kills empathy” is too strong. Classic research found reductions in perspective taking, compassion, and emotion recognition under some high-power conditions. Reviews also document lower social attention and greater distance. But the relationship is moderated by goals and construals.
Power can actually improve empathic accuracy when a prosocial orientation is active, because power helps people pursue the goal that is currently important. Responsibility-focused and entrusted-power studies similarly show more other-orientation under specific conditions.
The accurate conclusion is that power often reduces the need to attend to others, and therefore creates a risk of lower spontaneous social attunement. Whether that risk becomes actual empathic failure depends on motivation, role, context, and feedback.
What Psychology Can Say With Confidence
Several conclusions now have stronger support than the old slogan.
• Power reliably changes psychological functioning. The 2025 meta-analysis supports the broad approach–inhibition framework across a very large literature (Sabey et al., 2025).
• Power tends to facilitate action and goal pursuit rather than producing one fixed behavior (Galinsky, Gruenfeld, & Magee, 2003; Guinote, 2017).
• Power can reduce spontaneous perspective taking and social attention, but those effects are conditional rather than universal (Galinsky et al., 2006; Schmid Mast, Khademi, & Palese, 2020).
• Power can facilitate self-interest and unethical behavior, particularly under weak accountability, weak moral identity, or opportunity-focused construals (DeCelles et al., 2012; Pitesa & Thau, 2013; Foulk, Chighizola, & Chen, 2020).
• Power can also facilitate prosocial action when moral identity, responsibility, other-orientation, or prosocial goals are active (Scholl, 2020; Gobel, Choi, & Uchida, 2025; Hadar et al., 2026).
• Power, status, authority, personal control, dominance, and leadership are not interchangeable. Collapsing them produces misleading conclusions about who is powerful and what power does.
What Psychology Cannot Conclude
The evidence does not justify saying that all powerful people become immoral, that power permanently changes everyone’s personality, that reduced empathy is inevitable, that confidence proves narcissism, or that one laboratory manipulation reproduces the psychology of long-term institutional power.
It also cannot reduce abuse to the personality of the person at the top. Corruption can be enabled by networks, incentives, information bottlenecks, dependency structures, norms, and weak accountability. A psychologically complete account therefore includes both individual processes and the systems that make those processes consequential.
Finally, the direction of causality in real life can run both ways. Power changes behavior, and behavior changes who receives power, who keeps it, who is trusted, and who is challenged. Long-term power is a recursive social process rather than a one-time treatment.
How Organizations and Groups Can Reduce the Corrupting Risks of Power
No single intervention neutralizes power, and psychology does not offer a universal anti-corruption formula. Evidence does, however, support design principles that target known mechanisms.
• Preserve meaningful accountability. Require decision-makers to explain procedures and reasons, not merely outcomes, especially when they allocate resources that affect others (Pitesa & Thau, 2013).
• Protect independent feedback. Powerful people need information that does not depend on pleasing them.
• Reduce unnecessary social distance. Direct exposure to the people affected by decisions can restore information that hierarchical distance otherwise filters out.
• Separate personal autonomy from domination. Give people control over their work and lives without automatically giving them unchecked control over other people.
• Make responsibility explicit. Role expectations can frame power as stewardship of tasks and people rather than as personal entitlement (Scholl, 2020).
• Keep countervailing power real. Appeals, shared decision rights, independent review, and credible alternatives reduce one-sided dependence.
• Design for correction, not moral perfection. Even well-intentioned people become overconfident, inattentive, or narrowly goal-focused. Systems should make it possible to detect and reverse errors before they become durable patterns.
These principles follow from the same core insight as the research itself: power changes the social field. Good institutional design changes that field again by restoring information, dependence, scrutiny, and responsibility.
A Better Answer Than “Power Corrupts”
The old maxim survives because it captures a real danger. Power can reduce constraints, increase self-focus, widen social distance, amplify active goals, and make overconfidence more consequential. Those mechanisms can support corruption.
But the maxim becomes misleading when it is treated as a law of human nature. Modern research shows a more dynamic pattern. Power amplifies action; it can weaken attention to others; it can magnify moral identity or self-interest; it can invite opportunity-taking or responsibility; it can interact with status, legitimacy, accountability, and role expectations; and it can support costly prosocial action as well as exploitation.
The power paradox is therefore not simply that good people become bad. The deeper paradox is that influence can create psychological freedom from the very social signals that once made influence possible. Whether that freedom becomes corruption depends on what the person is pursuing and on whether the surrounding system still makes other people visible, consequential, and capable of answering back.
FAQ
Does power always corrupt?
No. Research supports conditional effects. Power can facilitate selfish, unethical, or exploitative behavior, but it can also strengthen prosocial or principled behavior when moral identity, responsibility, and prosocial goals are active.
Why can power corrupt people?
Power tends to increase approach motivation, action, confidence, and freedom from social constraint while sometimes reducing dependence on and attention to others. These changes can make self-serving goals easier to pursue and corrective feedback easier to ignore.
What is the power paradox?
Dacher Keltner’s power paradox describes the tension between gaining influence through socially valued behavior and then risking the loss of social sensitivity, restraint, or connection once power reduces dependence on other people. It is an integrative interpretation of research, not a universal law.
Does power reduce empathy?
It can reduce spontaneous perspective taking, compassion, and social attention in some contexts, but the effect is not inevitable. Prosocial goals, responsibility, and entrusted-power construals can preserve or increase other-orientation.
Does power cause narcissism?
No direct deterministic conclusion is supported. Power and narcissism are distinct. Some consequences of power can resemble narcissistic behavior, such as self-focus or overconfidence, but power is not a diagnosis and does not by itself establish a personality disorder.
Does power change personality?
Power changes psychological states and how motives are expressed. Evidence is stronger for effects on attention, affect, confidence, goal pursuit, and behavior than for an irreversible transformation of broad personality traits.
Can good people be corrupted by power?
Yes, situational mechanisms can affect people regardless of how they entered a role, but strong moral identity and responsibility can also make power support less self-interested behavior. Character matters, and so do institutions.
Does accountability stop power from corrupting?
Accountability can reduce self-serving behavior, especially when powerful decision-makers must justify the procedures they used. It is not a guarantee; the design, independence, and enforceability of accountability mechanisms matter.
Is power the same as status?
No. Power concerns control over valued outcomes; status concerns respect and esteem. Research increasingly treats them as distinct dimensions of hierarchy, and they can produce different behavioral effects.
Is power the same as authority?
No. Authority is a legitimate or recognized right to direct or decide. Power is broader and can come from resources, coercion, expertise, information, relationships, or dependence even without formal legitimacy.
What is the difference between power and personal control?
Power over others concerns the capacity to shape other people’s outcomes. Personal control concerns the capacity to shape one’s own life. Research suggests these can have different, even opposite, interpersonal consequences.
Related Articles
References
Cislak, A., Cichocka, A., Wojcik, A. D., & Frankowska, N. (2018). Power corrupts, but control does not: What stands behind the effects of holding high positions. Personality and Social Psychology Bulletin, 44(6), 944–957. https://doi.org/10.1177/0146167218757456
DeCelles, K. A., DeRue, D. S., Margolis, J. D., & Ceranic, T. L. (2012). Does power corrupt or enable? When and why power facilitates self-interested behavior. Journal of Applied Psychology, 97(3), 681–689. https://doi.org/10.1037/a0026811
Fast, N. J., Sivanathan, N., Mayer, N. D., & Galinsky, A. D. (2012). Power and overconfident decision-making. Organizational Behavior and Human Decision Processes, 117(2), 249–260. https://doi.org/10.1016/j.obhdp.2011.11.009
Fleischmann, A., & Lammers, J. (2020). Power and moral thinking. Current Opinion in Psychology, 33, 23–27. https://doi.org/10.1016/j.copsyc.2019.06.008
Foulk, T. A., Chighizola, N., & Chen, G. (2020). Power corrupts (or does it?): An examination of the boundary conditions of the antisocial effects of experienced power. Social and Personality Psychology Compass, 14, e12524. https://doi.org/10.1111/spc3.12524
Galinsky, A. D., Gruenfeld, D. H., & Magee, J. C. (2003). From power to action. Journal of Personality and Social Psychology, 85(3), 453–466. https://doi.org/10.1037/0022-3514.85.3.453
Galinsky, A. D., Magee, J. C., Inesi, M. E., & Gruenfeld, D. H. (2006). Power and perspectives not taken. Psychological Science, 17(12), 1068–1074. https://doi.org/10.1111/j.1467-9280.2006.01824.x
Giurge, L. M., van Dijke, M., Zheng, M. X., & De Cremer, D. (2021). Does power corrupt the mind? The influence of power on moral reasoning and self-interested behavior. The Leadership Quarterly, 32(4), 101288. https://doi.org/10.1016/j.leaqua.2019.03.003
Gobel, M. S., Choi, E., & Uchida, Y. (2025). Entrusted power enhances psychological other-orientation and altruistic behavioural tendencies. British Journal of Social Psychology, 64(2), e12857. https://doi.org/10.1111/bjso.12857
Gruenfeld, D. H., Inesi, M. E., Magee, J. C., & Galinsky, A. D. (2008). Power and the objectification of social targets. Journal of Personality and Social Psychology, 95(1), 111–127. https://doi.org/10.1037/0022-3514.95.1.111
Gu, Z., Liu, L., Tan, X., Liang, Y., Dang, J., Wei, C., Ren, D., Su, Q., & Wang, G. (2020). Does power corrupt? The moderating effect of status. International Journal of Psychology, 55(4), 499–508. https://doi.org/10.1002/ijop.12629
Guinote, A. (2017). How power affects people: Activating, wanting, and goal seeking. Annual Review of Psychology, 68, 353–381. https://doi.org/10.1146/annurev-psych-010416-044153
Hadar, B., Simchon, A., Gilead, M., & Smith, P. K. (2026). Elevated power promotes prosocial behavior more than elevated status. Personality and Social Psychology Bulletin, 52(7), 1796–1811. https://doi.org/10.1177/01461672251324678
Hu, M., Rucker, D. D., & Galinsky, A. D. (2016). From the immoral to the incorruptible: How prescriptive expectations turn the powerful into paragons of virtue. Personality and Social Psychology Bulletin, 42(6), 826–837. https://doi.org/10.1177/0146167216644428
Keltner, D. (2016). The power paradox: How we gain and lose influence. Penguin Press. https://www.penguinrandomhouse.com/books/312367/the-power-paradox-by-dacher-keltner/
Keltner, D., Gruenfeld, D. H., & Anderson, C. (2003). Power, approach, and inhibition. Psychological Review, 110(2), 265–284. https://doi.org/10.1037/0033-295X.110.2.265
Kipnis, D. (1972). Does power corrupt? Journal of Personality and Social Psychology, 24(1), 33–41. https://doi.org/10.1037/h0033390
Lammers, J., Stapel, D. A., & Galinsky, A. D. (2010). Power increases hypocrisy: Moralizing in reasoning, immorality in behavior. Psychological Science, 21(5), 737–744. https://doi.org/10.1177/0956797610368810
Magee, J. C. (2020). Power and social distance. Current Opinion in Psychology, 33, 33–37. https://doi.org/10.1016/j.copsyc.2019.06.005
Pitesa, M., & Thau, S. (2013). Masters of the universe: How power and accountability influence self-serving decisions under moral hazard. Journal of Applied Psychology, 98(3), 550–558. https://doi.org/10.1037/a0031697
Sabey, T. B., Shanklin, B. C., Colquitt, J. A., & Baer, M. D. (2025). The approach-inhibition theory of power: A meta-analytic test and synthesis. Psychological Bulletin, 151(10), 1245–1279. https://doi.org/10.1037/bul0000500
Schmid Mast, M., Khademi, M., & Palese, T. (2020). Power and social information processing. Current Opinion in Psychology, 33, 42–46. https://doi.org/10.1016/j.copsyc.2019.06.017
Scholl, A. (2020). Responsible power-holders: When and for what the powerful may assume responsibility. Current Opinion in Psychology, 33, 28–32. https://doi.org/10.1016/j.copsyc.2019.06.011
van Kleef, G. A., Oveis, C., van der Löwe, I., LuoKogan, A., Goetz, J., & Keltner, D. (2008). Power, distress, and compassion: Turning a blind eye to the suffering of others. Psychological Science, 19(12), 1315–1322. https://doi.org/10.1111/j.1467-9280.2008.02241.x