A violent minority? A quantitative analysis of those engaged in anti-immigrant violence in South Africa

ABSTRACT

Xenophobic violence poses a clear threat to South Africa, especially in light of the country’s treaty commitments to ‘open borders’ following the 30^th Ordinary Assembly of the African Union. Civil society organizations have been highly critical of government efforts to prevent xenophobic violence in the country. The aim of this study is to profile the determinants of individual participation in anti-immigrant violence. The article focused on both past and potential participation in this type of behaviour. My central thesis is that there is a robust correlation between involvement in peaceful organized anti-immigrant activities (e.g. boycotts and demonstrations) and participation in xenophobic violence. For this study, data from the 2015–2017 rounds of the South African Social Attitudes Survey (N = 9,292) was utilized. Bivariate and multivariate techniques were used to show that past and potential involvement in violent behaviour was strongly correlated with previous experience with peaceful anti-immigrant activity. By moving beyond a narrow focus on fixed episodes of collective violence, this article demonstrates the importance of a geographical analysis of xenophobic behaviour. However, there are limitations to the dataset used for this study and the article concludes by outlining future areas of research.

KEYWORDS:

• Xenophobia
• violence
• behaviour
• survey
• South Africa

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. The term refers to the use of physical force, intended to hurt or kill, with explicit purpose of preventing international immigrants from living or working in South Africa or any part thereof.

2. The exception to this general trend involves a highly publicized anti-immigration march in Tshwane organized by a group calling itself Mamelodi Concerned Residents (RCR). In February 2017 Mamelodi residents marched to the Home Affairs Department to hand over their memorandum asking that international migrants be removed from their community. Organizer Makgoka Lekganyane told the media that he and the RCR had organized a similar march to the Moffat View police station to submit the memorandum but felt that demonstration had attracted little attention for their cause (Whittles, 2017). The RCR’s February 2017 march seemed to spark anti-immigrant attacks in the area and the RCR denounced this violence as ‘hooliganism’.

3. Consider, for instance, the work of the Ad Hoc Joint Committee on Probing Violence Against Foreign Nationals. The committee was established by the South African Parliament to investigate the causes of collective violence against immigrants that had broken out in eThekwini and Johannesburg in April 2015. Instead of trying to track the kinds of people participating in the attacks, the committee was quick to deny that the attacks were not xenophobic. The committee’s final November 2015 report shifted blame for the attacks onto a range of external factors including the alleged unlawful activities of foreign shop-owners operating in the townships of eThekwini and Johannesburg (for the full report, see PMG, 2015).

4. For comprehensive reviews of this literature, see Neocosmos (2010), Matsinhe (2011), and Misago (2016).

5. Here the dataset was segmented between those communities which had been ‘hotspots’ during the largescale May 2008 attacks and those which had not. Crush et al. (2013) found that 9% of community members in such hotspots said that they would resort to violence to drive out irregular migrants. This can be compared to 11% of those living in non-hotspot communities.

6. Between January 2015 and January 2017, Xenowatch tracked a large number of xenophobic incidents. According to the platform, more than 500 shops were looted, 66 people died, 116 assaulted and 11,140 displaced due to xenophobic incidents over the period. Following this period there appeared to be a slowdown (with the exception of the North West) in xenophobic activity.

7. I explored how Economic Freedom Fighters (EFF) partisans self-reported past participation in anti-immigrant violence had altered during the 2015–2017 period. A substantial upsurge was observed in the portion of EFF supporters who told fieldworkers that they had not participated in this type of violence but might do so in the future between 2016 and 2017. The portion of EFF partisans making this claim increased from 7% in 2016 to 22% in 2017. A similar increase was not seen for supporters of the African National Congress or the Democratic Alliance.

8. The unanticipated nature of this outcome needs further explanation. I observed a significant increase in the portion of Indian and white individuals who admitted past involvement in violent behaviour during the period 2015–2016. Looking at the data more closely, it would appear that these minority groups in Gauteng had become more likely to admit participation in the distant past. It is not clear at this stage why we observe this finding. Social desirability bias of some sort could be at play here or it may be that the survey is picking up on a certain type of xenophobic violence that does not get reported on by the media. Further research is required to better understand this unexpected finding.

9. Consider, for instance, research by Cederman, Wimmer, and Min (2010) who looked at 157 different cases of ethnic violence during the period 1945 to 2005. They found strong statistical correlations between a group’s decline in status and the likelihood that a group would use violence against another group. In addition, consider research in the American South West by Olzak (1994) which showed a link between black economic and political gains and white-on-black attacks (also see Petersen, 2002 work in Eastern Europe).

Additional information

Funding

This work was supported by the Centre of Excellence in Human Development, University of the Witwatersrand [P20180003].