Petty corruption experiences and xenophobic violence in South Africa

ABSTRACT

Past studies have shown that corruption can promote conflict and instability. However, the capacity for corruption to influence anti-immigrant violence has received little academic attention. This article examines the relationship between experiences of petty corruption and public participation in violent xenophobic behaviour. It draws on Moral Foundations Theory to argue that corruption experiences undermine the moral inhibitors that prevent an individual from engaging in violence. The study focused on South Africa (where xenophobic violence seems to have accelerated during the last two decades) and used data from the South African Social Attitudes Survey. It found a robust association between the frequency of petty corruption experience and participation (past and intentional) in xenophobic violence. This finding holds even when controlling for variables that capture anti-immigrant sentiment, material deprivation and socio-political mobilisation. The article concludes by arguing that reducing xenophobic violence in South Africa will require strengthening efforts to decrease petty corruption.

KEYWORDS:

• Xenophobia
• corruption
• public opinion
• South Africa
• violence

Acknowledgement

For their support and encouragement, special thanks to Benjamin J. Roberts and Jarè Struwig Co-ordinators of the South African Social Attitudes Survey at the Human Sciences Research Council. Special thanks to the Faculty of Humanities at the University of Johannesbug where Steven Gordon is a Research Associate.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Xenowatch tracks violence using media reports, research publications, partner organisations, and original research as well as crowdsourcing (Misago et al., 2021). Because of the underreporting of hate crime by victims, relevant organisations and the media, the numbers produced by Xenowatch may dramatically undercount the actual level of xenophobic violence in the country.

2 There are many different definitions of corruption, but most involve the abuse of power for private gain. Indeed, as Bussell (2015) shows, there are significant (and fraught) definitional debates around how corruption should be conceived. No existing definition of corruption has managed to cover the many dimensions of the concept. Jain (2001) provides clarity to this issue, producing a typology that differentiates between different kinds of corruption. Three categories of corruption are identified: (i) grand; (ii) bureaucratic or petty; and (iii) legislative. Although not perfect, this typology provides necessary clarity for the purposes of this study.

3 For a discussion of the literature that looks at the multiple macro-level structural forces that have been put forward to explain xenophobic violence in post-apartheid South Africa, see Tewolde (2023).

4 Formal schooling was assessed using the question: ‘[w]hat is the highest level of education that you have ever completed?’ and was measured in terms of formal years of education completed.

5 When this question was asked as part of the SASAS 2006 round, only about a third of the adult population said that they (or their immediate family) had a corruption experience between 2001 and 2006. About a tenth (11%) of the general population told fieldworkers that this happened seldom, 13% occasionally, 6% quite often and 3% very often.

6 This approach was selected because of the nominal options offered to the respondents. In view of the multiplicity of decisional processes that were available to the respondents, linear or ordered traditional specifications would not be suitable. It was necessary to estimate error correlations across nominal choices and address the potential independence of irrelevant alternatives (IIA) assumptions. A multinomial probit model specification allows for comparisons between pairs of alternatives while not discounting the possibilities of IIA violations.