University of the Witwatersrand School of Social Sciences African Centre for Migration & Society ...

280 Pages·2016·2.07 MB·English

University of the Witwatersrand
School of Social Sciences
African Centre for Migration & Society

Migration, Governance and Violent Exclusion:
Exploring the Determinants of Xenophobic Violence in Post-Apartheid South Africa

Jean Pierre MISAGO
Student Number: 0404030D

Dissertation in fulfilment of the requirements for the degree of Doctor of Philosophy (PhD) in
Migration and Displacement Studies, School of Social Sciences, Faculty of Humanities,
University of the Witwatersrand

April 2016
Declaration
I declare that this thesis is my own unaided work. It is submitted for the degree of Doctor of
Philosophy in the University of the Witwatersrand, Johannesburg. It has not been submitted
for any other degree or examination in any other university.
Jean Pierre Misago

11 April 2016

ii
Dedication
To family members and all other victims of violent discrimination.

iii
Acknowledgements
I would like to express my most heartfelt gratitude to Prof. Loren Landau, my supervisor, for
the generous support throughout all my years of study and work at the University of the
Witwatersrand and for the guidance and patience that went far beyond supervision duties.
I am deeply indebted to the African Centre for Migration & Society (ACMS) for providing an
enabling and supportive environment.
I am also thankful for financial support received from the Social Science Research Council
(SSRC) and the Harry Frank Guggenheim Foundation.

iv
Table of Contents
I. INTRODUCTION ................................................................................................................................................ 1
1.1 SETTING THE SCENE.............................................................................................................................................. 1
1.2 XENOPHOBIC VIOLENCE: GLOBAL DIMENSIONS AND THEORETICAL APPROACHES ............................................. 9
1.3 STRUCTURE OF THESIS ....................................................................................................................................... 21
II. METHODS AND METHODOLOY: RESEARCHING THE DETEMINANTS OF XENOPHOBIC VIOLENCE .................... 30
2.1 INTRODUCTION .................................................................................................................................................. 30
2.2 XENOPHOBIA AND XENOPHOBIC VIOLENCE: CONCEPTUAL DISTINCTIONS AND CAUSAL LINKAGES ................ 31
2.3 THE NEED FOR A QUALITATIVE METHODOLOGICAL APPROACH AND A MICRO-ANALYTICAL FRAMEWORK ..... 40
2.4 DATA SOURCES AND THE USE OF SECONDARY ANALYSIS ................................................................................... 45
2.5 BASELINE STUDY: RESEARCH METHODS AND DESIGN ........................................................................................ 50
2.6 CONCLUSION ...................................................................................................................................................... 61
III. XENOPHOBIC VIOLENCE IN CONTEMPORARY SOUTH AFRICA: HISTORY, MORPHOLOGY AND POPULAR
EXPLANATIONS .............................................................................................................................................. 62
3.1 INTRODUCTION ................................................................................................................................................ 62
3.2 HISTORY AND NATURE OF XENOPHOBIC VIOLENCE IN CONTEMPORARY SOUTH AFRICA .................................. 63
3.3 POPULAR EXPLANATIONS AND THEIR LIMITATIONS .......................................................................................... 98
3.4 CONCLUSION .................................................................................................................................................... 118
IV. MICRO-POLITICS AND THE POLITICAL ECONOMY OF THE VIOLENCE: LOCAL DYNAMICS, ROLES AND
MOTIVATIONS OF VIOLENCE ENTREPRENEURS ............................................................................................ 121
4.1 INTRDODUCTION ............................................................................................................................................ 121
4.2 LOCAL DYNAMICS AND THEIR IMPACT ON XENOPHOBIC VIOLENCE ................................................................ 122
4.3 VIOLENCE ENTREPRENEURS’ ROLES AND MOTIVATIONS FOR VIOLENCE ......................................................... 146
4.4 CONCLUSION .................................................................................................................................................... 167
V. MOBILIZATION AS A TRIGGER OF XENOPHOBIC VIOLENCE ........................................................................... 170
5.1 INTRODUCTION .............................................................................................................................................. 170
5.2 PERSPECTIVES ON MOBILIZATION AS A KEY DETERMINANT OF COLLECTIVE VIOLENCE ................................. 172
5.3 MOBILIZATION TECHNIQUES AND PROCESSES FOR THE MAY 2006 XENOPHOBIC VIOLENCE ........................ 183
5.4 CONCLUSION: TOWARDS A ‘MOBILIZATION OF DISCONTENT’ MODEL .......................................................... 199
VI. LOCAL GOVERNANCE AND XENOPHOBIC VIOLENCE IN SOUTH AFRICA ........................................................ 203
6.1 INTRODUCTION .............................................................................................................................................. 203
6.2 LOCAL GOVERNANCE: A POLITICAL OPPORTUNITY STRUCTURE FOR XENOPHOBIC VIOLENCE IN SOUTH AFRICA
........................................................................................................................................................................ 204
6.3 ON SOCIAL CONTROLS AND COLLECTIVE VIOLENCE: THE STATUS QUO QUESTIONED.................................... 223
6.4 CONCLUSION .................................................................................................................................................. 234
VII. CONCLUSIONS: TOWARDS A GOVERNANCE MODEL OF XENOPHOBIC VIOLENCE ......................................... 237
7.1 XENOPHOBIC VIOLENCE AS ‘POLITICS BY OTHER MEANS’ .............................................................................. 238
7.2 THE TRIGGER EFFECT OF MOBILIZATION ........................................................................................................ 239
7.3 GOVERNANCE AND COLLECTIVE VIOLENCE: A MISUNDERSTOOD RELATIONSHIP .......................................... 242
7.4 A MULTIVARIATE MODEL OF XENOPHOBIC VIOLENCE ................................................................................... 245
7.5 FROM MULTIVARIATE TO THE GOVERNANCE MODEL OF XENOPHOBIC VIOLENCE ........................................ 251
7.6 CONCLUDING REMARKS ................................................................................................................................. 256
VIII. REFERENCES ................................................................................................................................................. 258
IX. ANNEX I: BASELINE STUDY RESEARCH THEMES ............................................................................................ 271

v
Abstract
Responding to inadequacies and limitations of current causal explanations for xenophobic
violence which has become a long standing feature in post-Apartheid South Africa, this study
proposes a Governance Model of Xenophobic Violence that provides a comprehensive
empirically-based and theoretically informed causal explanation. It is a multivariate empirical
and integrated theoretical explanatory model that identifies and explains the roles of - and the
complex interplays between - the key determinants of xenophobic violence consisting of
underlying causes, proximate factors and triggers. The six key determinants the model
identifies are: deprivation, xenophobic beliefs, collective discontent, political economy,
mobilization and governance. This study argues that these determinants and their
interconnections in a value-added process constitute the necessary and sufficient conditions
for the occurrence of xenophobic violence. I call it the governance model because of the
predominant role governance plays in the occurrence of xenophobic violence.
With underlying causes (deprivation, xenophobia and collective discontent) already
established, the study pays particular attention to the often missed proximate factors and
triggers (the political economy of the violence, mobilization and governance). It is through
the findings on these new factors that the study introduces new empirical and theoretical
insights and innovations to the understanding of, not only, xenophobic violence in South
Africa but also collective violence generally. First, this study argues that xenophobic violence
in South Africa is just ‘politics by other means and by doing so brings to the fore the often
missed centrality of micro-politics and localised political economy factors as key drivers of
collective violence particularly communal violence.
Second, the study argues that that the triggers of xenophobic violence and of collective
violence generally lie in the mobilization processes and not in the grievances and ensuing
discontent as argued by many theoretical approaches to collective violence. The study
suggests a new theoretical model, the Mobilization of Discontent Model, which captures the
increasingly recognised centrality of mobilization as a trigger of collective violence.
Third, the study argues that governance is a key determinant of xenophobic and collective
violence but not necessary in ways often assumed or prescribed by time-honoured and widely
accepted theoretical predictions, particularly those contending that collective violence and
other forms of contentious collective action tend to occur in societies where mechanisms of
social control have lost their restraining power. By demonstrating that local governance
deliberately facilitated the occurrence of xenophobic violence in areas where it occurred by
providing what I term micro-political opportunity structures, the study calls into question the
common understating of the relationship between governance and collective violence and
reveals some aspects of this relationship that are either misunderstood or undetected until
now.
The Governance Model of Xenophobic Violence this study proposes is an innovation that
clearly illustrates the poverty of most explanatory models of collective violence, which makes
it an appropriate tool for integrating empirical and theoretical knowledge from different
disciplines and for identifying gaps in existing scholarship.

vi
I. INTRODUCTION
XENOPHOBIC VIOLENCE: RETHINKING METHODOLOGICAL,
THEORETICAL AND EMPIRICAL FOUNDATIONS FOR CURRENT
UNDERSTANDINGS
1.1 SETTING THE SCENE
Xenophobic violence has become a longstanding feature in post-Apartheid South Africa.
Since 1994, tens of thousands of people have been harassed, attacked, or killed because of
their status as outsiders or foreign nationals. The violence was most intense and widely
scrutinised in May 2008. Indeed, the May 2008 violence, that started in Alexandra Township
on 11 May 2008 and quickly spread to other townships and informal settlements in the
country, was unprecedented in its intensity, ferocity and rapid geographic spread (Misago et
al, 2009).
Graphic images of violent attacks on foreign nationals (e.g. scenes of knife- and stick-
wielding aggressors, wounded victims, burning houses, and even a burning man) were seen
around the world. These scenes “were soon replaced by images of people who had fled in fear
of their lives to seek refuge in churches and police stations, eventually to be re-housed in tent
settlements like those housing famine or war refugees” (Dodson, 2010:3). In less than a
month, 135 separate violence incidents (Bekker et al, 2008) resulted in at least 62 people
dead; 670 wounded; dozens raped; more than 100 000 displaced; and millions of Rand worth
of property looted, destroyed or appropriated by local residents (CoRMSA, 2008). Most
victims were African foreign nationals but a ‘third were South Africans who had married
foreigners, refused to participate in the violent orgy, or had the misfortune of belonging to

1
groups that were evidently not South African enough to claim their patch of urban space”
(Landau, 2011:1).
After initially denying the crisis and offering unheeded appeals for calm, the government
deployed the army forces to contain the violence in June 2008 (Ibid). However, while the
violence subsided with the deployment of the army, it did not end there. Indeed, despite
government’s claims to the contrary (Black Sash, 2009), violence against foreigners continued
post-May 2008 and was increasingly reaching ‘disturbing proportions’ (DAC, 2012). Attacks
on non-nationals continue to be regularly reported resulting in rising cases of murder, injuries,
threats of mob violence, looting and the destruction of residential property and businesses, as
1
well as mass displacement (CoRMSA, 2011; UNHCR ROSA, 2013 ; UNHCR ROSA, 2015) .
Perhaps not surprisingly, t he unprecedented nature of the May 2008 violence triggered a
frenzy of analyses and commentary as scholars , policy analysts and government officials
attempted to make sense of what was happening in the mu ltiracial ‘rainbow’ nation (Fauvelle-
Aymar et al., 2011; see also Nieftagodien, 2011). The initial official denialism and apparent
public surprise (despite many and clear warning signs) were immediately followed by
reactionary attempts to explain the causes of the violence. Initial reactions were
understandably speculative and based on inaccurate or outdated information, political
rationales and ideological stances rather than empirical evidence (Fauvelle-Aymar et al,
2011). As such, these explanations (that included ‘afrophobia’, ‘third force’, poor border
control, poor service delivery, poverty, unemployment, corruption - see for example HSRC,
2008; Landau and Misago, 2009) failed to take into account many underlying historical and
socio-political causes as well as specific factors that triggered violence in some areas and not
in others.

1
Unpublished UNHCR stats on data compiled from the UNHCR xenophobia hotline and Communique by UNHCR’s
Munyaneza Alphonse at the Protection Working Group meeting of 22 January 2013.

2
Realising that initial reactions and explanations did not adequately explain the occurrence of
violence, discussions and debates (among scholars, political actors and practitioners)
continued over the root and immediate causes of the violence, the effectiveness of the
responses provided and appropriate measures to be put in place to prevent future occurrences.
Debates have led to several competing causal explanations that detail broad economic,
political, historical, structural and attitudinal factors (see for example Crush et al, 2008;
Neocosmos, 2008; Sharp, 2008; Pillay, 2008 and Dodson, 2010). While valuable in providing
the historical, political and socio-economic context in which the violence occurred, these
explanations similarly “falter when faced with empirical or logical interrogation” (Landau
2010:215) as they fail to account for i) the appearance of violence in some areas while others
with similar socio-economic conditions remained calm; ii) the specific targeting of certain
groups of foreign nationals and the murders of so many South African citizens and iii) the
timing, location and diffusion of the violence.
Indeed, as this study clearly demonstrates, existing explanations evidently maintain only a
tenuous relationship with empirical reality (with regard to the occurrence of the violence) and
actually reveal a methodological and analytical ‘laziness’ - or at least a lack of analytical
rigour - characteristic of hurried analysts who impose a priori explanations to current
phenomena without substantive or empirical backing (Horowitz, 2001). Disappointingly,
these explanations rarely attempt to answer those ‘deceptively’ simple but critical questions
that illuminate the occurrence of the violence: how, when and where; ‘why here and not
there’; ‘why now and not then’? (See details on these questions in Horowitz, 2001).
Analytically, this thesis argues that the poverty of existing explanatory models lies in a
twofold repertoire of analytical blind spots: i) the lack of empirical backing as indicated
above; and ii) incompleteness as most offer reductionist, one-factor explanations for such a
complex social phenomenon and as such can be at best partial or incomplete. Perhaps most

3
importantly, they all suffer from blindness to their own shortcomings particularly with mono-
causal explanations claiming to be all encompassing (i.e. identifying only one factor among
many but invalidly claiming to explain the entire process leading to the occurrence of the
violence thereby claiming to account for all the elements of the violence causal chain).
Further, a review of the literature reveals that most of the existing research, by virtue of its
focus on establishing (if unsuccessfully) ‘Why’ xenophobic violence occurred, has neglected
those questions that are critical in understanding the occurrence of the violence. The ‘How’
question is particularly important if we want to identify and understand processes through
which conditions and attitudes lead to or translate into mass violent attacks. Few attempts
have been made to establish the link between those general structural factors or even the
pervasive xenophobic sentiments and the outbreak of the violence. Only by answering the
‘why’ and ‘how’ questions are we able to understand and identify all the key determinants (i.e.
the underlying factors, and proximate causes and the triggers) of the violence. Similarly, even
fewer efforts have been made to explore xenophobic violence and its determinants in relation
to established conceptual and theoretical frameworks (Crawely, 2011).
While acknowledging that, since 2008, the violence has received increasing scholarly
attention, this limited nature of current explanatory models call for i) a reassessment of our
current understanding of the phenomenon and the empirical, theoretical, and methodological
assumptions on which it is based, and ii) further academic enquiries that move beyond
generalisations, provide theoretically informed and empirically based explanations. There is
indeed a recognition that “No one has yet fully explained either the prevalence of a culture of
xenophobia in South Africa or the particular outbreak of that culture into specific acts of
brutality in May 2008” (Dodson, 2010:11).

4