How Power Affects Influence

Power and Influence in a Nutshell

Is it easier or more difficult to influence people when they feel powerful? In general, power acts like an amplifier—it strengthens existing thoughts and attitudes, and it also drives people into action. So when you want to change people’s minds, it is best to try convincing people when they feel less powerful. Yet, if people already agree with you and you only want them to take action, then they are more likely to do so when they feel powerful.

Power Tends to Amplify, and Absolute Power Amplifies Absolutely

An English politician and historian Lord Acton famously said that “power tends to corrupt, and absolute power corrupts absolutely.” Acton also added that “great men are almost always bad men.” Such is the popular understanding of what power does to people. Acton was no doubt on to something, but it seems that his great observation was not absolutely correct.

Power is a complicated social phenomenon, and social psychologists still don’t know a good deal about its main effects and its many side effects. Yet, research shows, for one, that power works like an amplifier: it automates people’s thinking, for example by encouraging greater reliance on existing attitudes and stereotypes, and it also drives people into action. A 2001 study by Serena Chen, Annette Lee-Chai, and John Bargh found, for example, that people with charitable and pro-social bent are more likely to translate their unselfish and altruistic motivations into action when they feel powerful. Conversely, research suggests that it might be not the smartest idea to give bad people more power.

Power and Persuadability

Since power operates like an amplifier, it strengthens people’s confidence in their existing attitudes. So power makes it more difficult to change people’s minds. This is good news if people’s existing attitudes are favorable, and this bad news if their attitudes are unfavorable. Powerful people also process information less carefully and systemically, so they automate their judgments. This is good news if you want people to make fast, heuristic decisions, but this is bad news if you want to them to think carefully and analytically.

Research on Power on Persuadability

A 2007 study by Pablo Brinol, Richard Petty and their colleagues, tested how power would affect persuadability. Some of their participants were primed with high-power: they engaged in a role-playing exercise where they played a manager. The manager was told that he or she completely controlled the work process, could freely evaluate the subordinates, and allocate the rewards. The manager also got to sit in a taller and better-looking chair than the subordinate. People in the low-power condition played the subordinates; they were told they had no control over how the work was done, over the evaluation, or the division of resources.

All the participants then had to evaluate an ad for a new cell phone. Compared with low-power participants, high-power participants were much less swayed by the ad with strong arguments. This shows that the low-power people thought about the reasons more carefully, and high-power participants used heuristic assessment.

Practical Application

The main takeaway here is to avoid occasions when people feel powerful if you want them to consider your ideas more carefully and systemically or whenever their existing attitudes, ideas, or thoughts are likely to be unfavorable. Conversely, it might be useful to try convincing people when they feel powerful if you want them to make simple, heuristic decisions or if they already agree with you and you only want them to take some action.


  1. Serena S. Chen, Annette Y. Lee-Chai, and John A. Bargh, Relationship Orientation as Moderator of the Effects of Social Power,  Journal of Personality and Social Psychology, 80, 183-187 (2001).
  2. Pablo Brinol, Richard E. Petty, Carmen Valle, Derek D. Rucker, Alberto Becerra, The Effects of Message Recipients’ Power Before and After Persuasion: A Self-Validation Analysis, Journal of Personality and Social Psychology, 93, 1040–1053 (2007).