Finding Partisanship Where We Least Expect it: Evidence of Partisan Bias in a New African Democracy

Elizabeth Carlson

Research output: Contribution to journalArticlepeer-review

32 Scopus citations


Much of the literature on political behavior in Africa’s new semi-democracies has treated partisan affiliation as weak, purely pragmatic, or a proxy for other, more meaningful identities such as ethnicity. In this article, I dispute these conceptions by demonstrating that partisanship in an African context, like partisanship in established democracies, is a psychologically meaningful identity that can inspire voters to engage in motivated reasoning. By combining survey data with an original dataset of objective indicators of local public goods quality in Uganda, I show that supporters of the incumbent president systematically overestimate what they have received from government, while opposition supporters systematically underestimate. Partisan support precedes, rather than results from, this mis-estimation. I also show that partisans of the incumbent (opposition) are significantly more (less) likely to attribute any bad outcomes they observe to private actors rather than the government. I argue that these findings are consistent with the predictions of social identity theory: the conflict that marked many African political transitions, and the mapping of African parties onto existing social cleavages, are sufficient conditions for the creation of strong political-social identities like those that characterize partisanship in the West. My findings indicate that Africanists should take partisanship seriously as a predictor of political behavior and attitudes.

Original languageEnglish (US)
Pages (from-to)129-154
Number of pages26
JournalPolitical Behavior
Issue number1
StatePublished - Mar 1 2016

All Science Journal Classification (ASJC) codes

  • Sociology and Political Science


Dive into the research topics of 'Finding Partisanship Where We Least Expect it: Evidence of Partisan Bias in a New African Democracy'. Together they form a unique fingerprint.

Cite this