2020/10/06 by Louise Bøttkjær, Mogens K. Justesen · 25 citations
Social Sciences · Economics, Econometrics and Finance · #Corruption and Economic Development #Taxation and Compliance Studies #Culture, Economy, and Development Studies #Clientelism #Language change #Politics #Political science #Democracy #Political economy #Poverty #Corrupt practices #Public office #Work (physics) #Political corruption #Economics #Law
paper · open access · doi:10.1086/710146
published in The Journal of Politics 83(2), 788-793 (University of Chicago Press)
openalex publication_date 2020/10/06 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/01
Democratic elections are supposed to prevent corrupt politicians from winning office. In practice, however, corrupt politicians frequently get elected for public office in mass elections. In this article, we examine how voters respond to corruption allegations against political candidates and how voter responses may be attenuated by patronage and partisanship. We test these explanations through a survey experiment in South Africa—a country where political corruption is a recurrent issue. We find that voters express strong willingness to punish corrupt candidates. However, voters are more lenient toward corrupt politicians when they are offered work or jobs in return for their vote as part of a clientelist exchange. This effect is amplified for people living in economic hardship. These findings suggest that some types of clientelism serve to reproduce corruption—particularly when poverty is rife—and have important implications for the fight against corruption.