Identity and inequality misperceptions, demographic determinants and efficacy of corrective measures

K. Peren Arin, Deni Mazrekaj, Marcel Thum*, Juan A. Lacomba, Francisco Lagos

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

By conducting two waves of large-scale surveys in the United Kingdom and Germany, we investigate the determinants of identity and inequality misperceptions. We first show that people substantially overestimate the share of immigrants, Muslims, people under the poverty line, and the income share of the richest. Moreover, women, lower-income, and lower-educated respondents generally have higher misperceptions. Only income share misperceptions are associated more with people who place themselves on the left of the political spectrum. In contrast, the other three misperceptions are more prevalent among those who place themselves to the right. We then attempt to correct misperceptions by conducting a classic controlled experiment. Specifically, we randomly assign respondents into a treatment group informed about their initial misperceptions and a control group left uninformed. Our results indicate that information treatments had some corrective effects on misperceptions in Germany but were ineffective in the United Kingdom. Moreover, information treatments in Germany were more effective for men, centrists, and highly educated respondents. There is also no evidence of spill-over effects: correcting one misperception does not have corrective effects for the other misperceptions.
Original languageEnglish
Article number12300
JournalScientific Reports
Volume14
Issue number1
DOIs
Publication statusPublished - 29 May 2024

Keywords

  • COVID-19
  • Identity
  • Immigration
  • Inequality
  • Information treatment
  • Misinformation
  • Perception bias
  • Poverty

Fingerprint

Dive into the research topics of 'Identity and inequality misperceptions, demographic determinants and efficacy of corrective measures'. Together they form a unique fingerprint.

Cite this