The relationship between power and secrecy

Shane Schweitzer*, Rachel Lise Ruttan, Adam Waytz

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

1 Scopus citations


Seven studies test and find evidence for a relationship between secrecy and power. People who received secret information from another person felt more powerful than people who did not, both in terms of secrets they recalled from life (Studies 1, 2, and 4) as well as secrets from strangers (Studies 3 and 5). The effect of receiving secret information on experiencing increased power is not driven by changes in affect, nor is it contingent on the secret offering any instrumental advantage to the self. We test potential psychological mechanisms across three studies, specifically, exclusivity of the secret (Study 2), whether receiving a secret increases people's sense of felt reliance (Study 3), and whether receiving a secret makes people feel more trusted (Study 4). We find the effect is not dependent on the exclusivity of the secret, but that both a sense of reliance and feeling trusted drive the relationship between receiving a secret and feeling powerful. We also find that receiving a secret not only increases power, but also has downstream consequences in terms of increasing illusory control over the secret-giver and over others (Study 5). Studies 6 and 7 test and show that power reduces people's willingness to share secret information. People induced to feel powerful were less likely to share secrets about an organization with others (Study 6) and were more likely to approve of non-compete agreements in employment contracts (Study 7).

Original languageEnglish (US)
Article number104300
JournalJournal of Experimental Social Psychology
StatePublished - May 2022


  • Power
  • Reliance
  • Secrecy
  • Trust

ASJC Scopus subject areas

  • Social Psychology
  • Sociology and Political Science


Dive into the research topics of 'The relationship between power and secrecy'. Together they form a unique fingerprint.

Cite this