SOCIAL DESIRABILITY BIAS AND REPORTED VOTE PREFERENCES IN AFRICAN SURVEYS

Report an error

Summary

AI-generated

This summary is written by a language model reading the source document. It is not the publisher's words and is not a substitute for the original.

Learn more about AI enrichment

The paper argues that African voters intentionally hide ethnic preferences in surveys to avoid appearing 'tribalistic'. Experimental evidence from Uganda shows that voters are significantly less likely to choose a coethnic candidate when reporting their choice aloud compared to using a secret ballot. Analysis of Afrobarometer data suggests that observation by others during interviews reduces the likelihood of reporting ethnic preferences by approximately 10%. The author concludes that this bias may lead researchers to underestimate the role of ethnicity in voting and overestimate the importance of candidate performance.

Key insights

AI-generated

These insights are written by a language model reading the source document. They are not the publisher's words and are not a substitute for the original.

Learn more about AI enrichment
  • In a Ugandan voting simulation, respondents casting secret ballots viewed a candidate's ethnicity as four times more important than their performance; coethnicity increased a candidate's likelihood of winning by 12%, while a record of provision increased it by only 3%.
  • When required to report their vote aloud in the Ugandan experiment, voters' behavior reversed: the impact of ethnicity became insignificant, and they instead prioritized a candidate's record of goods provision, which increased the likelihood of winning by 12%.
  • The reduction in coethnic support during public voting was most pronounced for candidates without a record of provision; their victory rate dropped from 47% (secret) to 37% (public), whereas candidates with a record saw only a 4% decline.
  • Analysis of Afrobarometer data across multiple countries indicates that being observed by others during a survey interview reduces the likelihood of a respondent reporting a preference for a coethnic candidate by approximately 10%.
  • In Burkina Faso, Senegal, and Madagascar, the impact of observation was particularly large, with respondents being 10% to 15% less likely to report coethnic support; notably, these are countries where ethnicity was previously thought to be a poor predictor of voting.
  • Coethnics of an incumbent are 11% less likely to report any electoral preference at all compared to non-coethnics, suggesting they may use non-response to hide ethnic preferences.
  • Simulation analysis suggests that if approximately 16% of coethnic voters who have a true preference for an incumbent either report a vote for the opposition or fail to report a preference, the estimated impact of coethnicity on the vote becomes significantly attenuated.

Cite the original document

APA
Carlson, E. (2014). SOCIAL DESIRABILITY BIAS AND REPORTED VOTE PREFERENCES IN AFRICAN SURVEYS. Afrobarometer. https://www.afrobarometer.org/wp-content/uploads/2022/02/Afropaperno144.pdf
Chicago
Carlson, Elizabeth. SOCIAL DESIRABILITY BIAS AND REPORTED VOTE PREFERENCES IN AFRICAN SURVEYS. Afrobarometer, 2014. https://www.afrobarometer.org/wp-content/uploads/2022/02/Afropaperno144.pdf.
Wikipedia
{{cite report |last1=Carlson |first1=Elizabeth |title=SOCIAL DESIRABILITY BIAS AND REPORTED VOTE PREFERENCES IN AFRICAN SURVEYS |publisher=Afrobarometer |date=January 2014 |url=https://www.afrobarometer.org/wp-content/uploads/2022/02/Afropaperno144.pdf |access-date=5 November 2026 |via=Climate Insights Directory (document afbarom-000505)}}
BibTeX
@techreport{carlson2014social, author = {Carlson, Elizabeth}, title = {{SOCIAL DESIRABILITY BIAS AND REPORTED VOTE PREFERENCES IN AFRICAN SURVEYS}}, institution = {Afrobarometer}, year = {2014}, month = jan, url = {https://www.afrobarometer.org/wp-content/uploads/2022/02/Afropaperno144.pdf}, urldate = {2026-11-05}, note = {Indexed by Climate Insights Directory, document afbarom-000505} }

Full text

Collected · Record updated