Is South Africa Reverting to a Repressive State?

Since the Marikana  massacre in 2012, several journalists, academics and media commentators have argued that South Africa is reverting to a repressive state. They have interpreted violence at the hands of the South African Police Service (SAPS) generally, and Marikana specifically, as signs that the post-apartheid social order can no longer be held in check through consent alone. They argue that the ruling African National Congress (ANC) and other powerful actors have concluded that naked violence is now needed to stabilise increasingly fractious social relations.1 Some have even used the term “police state” to describe post-Marikana South Africa.2  As a police state is one where the police act as a political force to contain social dissent using arbitrary force, it is an important manifestation of a more repressive state.

How likely is South Africa to descend into  a state of full-blown repression? How likely is it that there will be more Marikanas? Need- less to say, being able to answer these questions will have a major impact on the future trajectory of the  country’s politics. There can  be little  argument with  the  statement that South Africa’s democratic government under its  fourth president, Jacob  Zuma, has strengthened the coercive capacities of the state, consisting of the police, the intelligence and  the  military and  located in the Justice, Crime Prevention and Security Cluster.  In fact, it would appear that this  Cluster  has  become the  praetorian guards of an increasingly embattled presidency.3 The well-reported growth in the levels of police violence against ordinary civilians and  protestors and  police militarisation are the most visible manifestation of this shift, as is the normalisation of the military in domes- tic  policing functions, which suggests a growing militarisation of society.4 However, the  huge  public controversies over  police violence and  police militarisation, mask the  fact  that there are  fundamental shifts in the coercive capacities of the state, away from  overt  repression and  towards less visible, more pre-emptive forms of repression. What are the indicators of this shift and why is it significant?

Members of the Association of Mineworkers and Construction Union (AMCU) dance and sing around Wonderkop Hill during the 4th anniversary of the Marikana shooting in Rustenburg, South Africa, 16 August 2016. Photo: ANP/EPA/Kim Ludbrook

From Human Intelligence to Signals Intelligence

The first indicator is that intelligence work has become increasingly important to stabilising social relations. Surveillance provides the state with a politically low-cost form  of social  control, as abuses are  very difficult to detect. Political surveillance is part of an arsenal of tools available to the state to pro- file problem subjects, and to use this knowledge  to stymie protests they  may  consider to be problematic. The  state can  use  such surveillance, or the  threat of surveillance, to create fear that organised violence will be used against perceived opponents.

At the  same time,   the  fear  of  being watched may  force  people to  self-police their own behaviour.

In South Africa, the state has expanded its surveillance capacities over the past decade.  In 2003, the  Thabo Mbeki  Presidency issued a directive requiring an  expansion of the-then National Intelligence Agency’s (NIA) mandate to include political and  economic intelligence. In the  case  of political intelligence, the NIA was to focus on “…the strengths and  the  weaknesses of political formations, their constitutions and  plans, political figures  and  their roles  in governance, etc”.5 These  changes led to the  intelligence services ballooning in size.  A year later, signs emerged that intelligence operatives were  becoming embroiled in internal factional battles in the ANC: a problem that was  proved to  exist  by  a  Commission of Enquiry, which partly blamed the  culture of secrecy in the  intelligence services as a source of the problem.6

Shortly  after  Zuma  took  office,  the domestic and  foreign intelligence services were  centralised into  the  State  Security Agency  (SSA). The  political intelligence- gathering mandate has also allowed the government to normalise spying on  domestic political groupings on the  most tenuous of grounds.7

A document leaked to the media, and  apparently summarising the  SSA’s National Intelligence Priorities for  2014 (which are classified, although they  should not  be)  – and  which are  developed every year  to guide  the  use  of the  state’s surveillance capacities – states that the SSA should investigate and  engage in counter-planning for a “so-called “Arab Spring”  uprising prior to [2014 national] elections’.8 The SSA claims it will resort to the “maximum use of covert human and  technical means” to  counter these threats.9 The document’s citing  of the Arab spring – a legitimate struggle against authoritarianism – is significant, as it implies that this protest wave in the Arab region was essentially illegitimate. In the South African context, the  risk of such a priority straying from the covert surveillance of illegal political activity  into  legitimate activities should be self-evident: a risk that is strengthened by the SSA’s overly broad mandate, excessive secrecy, recent history of abuse of this man- date and  inadequate reforms to  increase public accountability.

South African Police fire stand guard during a protest by students supporting the “Fees Must Fall movement” outside the South African Parliament where South African Minster of Finance delivers his Mid-term Budget address in Cape Town, on October 26, 2016. Photo: ANP/AFP / Schalk van Zuydam

The  risk associated with  human intelligence is that the  identities of intelligence operatives deployed to  spy  on  organisations can  always  be  uncovered, leading to  politically-costly scandals about intelligence abuses. As a result, the  intelligence community has  taken advantage of  the digital “revolution” to shift away from using human intelligence (intelligence gathered through physical means) to  signals intelligence (intelligence gathered from  communications surveillance). It is difficult to tell whether South Africa has embraced this global  shift,  but  it would be  unsurprising if it has.  While  the  government’s targeted interception  capacities are  regulated in terms of the  Regulation of Interception of Communications and Provision of Communications related Information Act (Rica), mass  surveillance remains  completely unregulated in terms of the law, which pre- disposes these capacities to abuse. In fact, not  only  does  South Africa produce mass surveillance technology, but  the  state has funded its  development10  and  allowed it to be exported, including to authoritarian regimes such as Libya, where the equipment was  used to  spy  on  Muammar Gaddafi’s political opponents.11

From Militarised Policing to Intelligence-led Policing

The second indicator, closely  related to the first, is the shift from militarised policing to intelligence-led policing. As its name suggests,  this  policing model uses  risk assessment as  its  main tool  to  direct policing decisions about where and  how  to  intervene. The model is more recent than para- military policing, as it was conceptualised in Britain and  the United States (US) in the 1990’s, but  it really  gained currency after the  September 11 attacks on New York and Washington. Intelligence-led policing relies heavily  on  covert techniques for  crime- detection, including paying informants, spying  on  individuals and  organisations, the use of Closed Circuit Television (CCTV ) cameras, communications surveillance and the interception of voice and  data traffic.12

Intelligence-led policing does  not  necessarily make human rights violations go away;  it merely makes them less  visible. This form  of policing encourages problematic profiling of individuals or social groups that may resort to crime, which can  lead  to stereotyping of particular social  groups as being predisposed to crime. Activists  who are  considered to  be  politically threatening to existing ruling groups may be placed under surveillance to gain  more information  about their activities and  to intimidate them, which risks  chilling political activity. The US police used intelligence-driven policing to infiltrate organisations linked to the  anti-globalisation movement, to identify and  isolate “troublemakers”.13  But like overt forms of violence, generalised surveillance techniques also erode public trust in the  state: in fact, the  latter can  do so more readily than the former as surveillance proceeds from  the  premise that states do not trust their citizens from the outset.

Intelligence-led policing relies heavily on covert techniques for crime- detection, including paying informants, spying on individuals and organisations, the use of Closed Circuit Television (CCTV ) cameras Photo: Centralhighlandssecurity

In South Africa, the  Crime  Intelligence Division of SAPS holds the  key to this  new policing strategy, so it is unsurprising that this Division has become so powerful (and controversial) in recent years,  as this policing model makes it the  lynchpin of policing strategies.14 Heightened power without heightened accountability is a recipe for disaster. A case currently being heard in the Pretoria Commercial Crimes Court points to members in the division having used the surveillance capacities of the  state to spy on journalists. Yet, in spite of its increasing importance to policing work, there are signs of Crime  Intelligence having lost its effectiveness, leading to a resurgence of organised crime.15

SAPS has  embraced intelligence-led policing for several reasons. Police violence is  eroding trust between the  police and communities, making it more difficult to revert back  to community policing.16 Yet at the same time, SAPS cannot risk many more high  profile shoot-outs with  protestors, as the  long-term political costs  will simply  be too  great. So, it stands to  reason that the SAPS would search for  a  policing model that still allowed them to contain dissent using a less politically-risky approach, and intelligence-led policing provides just such a model.

From Post-hoc to Preemptive Repression of Protests

The third indicator is an  increasing use  of pre-emptive methods of containing protests through manipulations of the Regulation of Gatherings Act (RGA), to stop more protests from spilling  out onto the streets in the first place. In a research study I led on the  right to protest in eleven municipalities17  – and which involved the  physical collection and logging of municipal data about gatherings and  protests over a five year period (2008-13) – I found that none of the  municipalities studied received a clean bill of health. A research team collected all notifications for protests and  gatherings sent  to municipalities in terms of the RGA: they yielded incredibly rich data about how many protests were taking  place  relative to gatherings, the  rea- sons  for the protests, the protest actors and municipal responses to the protests.

Police officers arrest a Congolese protester after a protest against the Democratic Republic of the Congo President, Joseph Kabila outside the DRC’s Embassy on December 20, 2016 in Pretoria. Photo: ANP / AFP John Wessels

The municipal and  the  police statistics suggest that the  majority of protests take place  peacefully and  uneventfully, which is not  the  dominant image of protests either in  the  media or  the  public imagination. In fact,  from  SAPS’s Incident Registration Information System (IRIS) database for the areas with the most unrest-related incidents between 2009 and  2012 18, it became clear that despite being labelled unrest-related, most of the protests did not escalate beyond barricade-building and  tyre-burning, and in fact  were  recorded as being fairly incident-free.19  The  protests recorded in  the municipal records constitute a humdrum of protests, taking  place  day  in,  day  out throughout the country with little incident. Between the  media and  police hype  about “violent service  delivery protests’, it is this wider  picture of peaceful protests that is so often missed, and  unsurprisingly so. The security cluster can  use images of marauding mobs, apparently predisposed to violence, to create moral panics in the  public about protests, to turn the  public against protestors (even  those whose demands are legitimate), and  to justify heightened security measures against them.

Yet in spite of protests remaining largely peaceful, all the  municipalities surveyed instituted unreasonable restrictions on the right  to  protest, and  these have  curtailed this right to varying  degrees. While the misapplication of the RGA has been a problem at least  since  the  early  2000’s20, a particularly significant shift became apparent from 2012 onwards. In the wake of the local government elections, the  Department of Co- operative Governance sent  out a circular to local  governments that outlined proactive measures that municipalities need to take to deal with protests. These measures included “… [working] with the office of the speaker [and]  public participation units to ensure ongoing engagement between councillors and  communities and  residents’.21  Several municipalities used this memo as a pretext to change how they administered the RGA.

This shift increased the already-onerous bureaucratic obstacles municipalities put on protests, many of which already shared an assumption that the notification process in terms of the  RGA was actually a permission-seeking exercise, and that they had the right  to grant or deny  “permission” to convenors to engage in a gathering or protest. This municipal misapprehension of the pro- cess set the tone for how notifications were dealt with,  both by the  municipalities and by the police. Practices that limited the right unduly included a requirement on the part of convenors to seek a letter from the institution  or person they were marching against, guaranteeing that they  would be willing to accept the memorandum. The rationale for seeking such an assurance appears to be to prevent frustration on the part of protestors, which could boil over  into  violence. How- ever, it has also become a censorship device, where those who are being marched against can squash the protest simply by refusing to accept the memorandum.

A student wears a T-Shirt with the defaced image of president Jacob Zuma over his head during a student protest over fees to parliament on the occasion of South African Finance Minister Pravin Gordhan presenting his Medium Term Budget Policy speech to Parliament in Cape Town, South Africa, 26 October 2016. Police used stun grenades and rubber bullets to disperse student protestors demonstrating at the gates of parliament as the Finance Minister delivered his medium term budget policy speech. Photo: ANP/ EPA/NIC BOTHMA

The City of Johannesburg requires protest  convenors to seek  permission from  a ward  councillor to  protest, and  after  the 2012 Co-operative Governance memo, they and  the Mbombela municipality, instituted a filtering system to reduce the  number of service  delivery protests, where convenors need to show  that a meeting took place between the  mayor’s office,  the  community  and  the  ward  councillor involved in that community, or at the very least  that an attempt was made to bring  all parties to the table to resolve the issues at hand.22 But this prescription is not  lawful,  as the  RGA does not  prescribe what  process people should follow  before they  take  to the  streets.

While the municipalities studied have gradually closed spaces for the right to protest, this closure is highly uneven and subject to considerable contestation. Spaces were much more closed where the political and economic elite were united in their intentions to stifle protests.

The number of “approved” protests increased in  Mbombela once the  filtering process was introduced, suggesting that the  potentially “troublesome” protests did  not  even enter into  the system. But the municipality did admit that the  condition had  led to an increase in the  number of “unrest-related” protests, taking place outside the framework of the  RGA, and  that the  police were  more likely to  be  heavy-handed against such a protest. These  were led mainly by individuals or organisations that were in dispute with the structures they were meant to negotiate with, suggesting that an increasingly restrictive approach towards protests on the part of the municipality was changing the character of the protests, forcing them to become what the  authorities would consider “unlawful” and driving up the potential for the protests to become disruptive.

While  the  municipalities studied have gradually closed spaces for the right to protest,  this  closure is highly  uneven and  subject  to  considerable contestation. Spaces were  much more closed where the  political and  economic elite were united in their intentions to  stifle  protests and  prevent criticism and  alternative forms of mobilisation: but  this  unity was  not  found uniformly  across the  municipalities. As Oliver has  argued, erratic government repression arises not because the government has chosen to be erratic, but because of inconsistencies among political actors.23 Furthermore, non-conventional actors are more likely to be repressed than conventional ones (such as unions or well-known political parties), as the  security apparatus consider the  former  to be less predictable than the  latter.24

University of Johannesburg students raise their hands while blocking a motorway during clashes with riot police on September 28, 2016 outside the University of Johannesburg Campus.
The issue of education fees ignited widespread frustration over a lack of opportunities for young South Africans, worsened by a weakening economy and high unemployment. Photo: ANP/AFP. Marco Longari

The  evidence supports a view of the  state put  forward by Gramsci that it is not monolithic,  but  is rather a site where ruling  class alliances take place  or even  shift.25 In times of significant political de-alignment, elements of the state can even work against one another. Erratic repression is likely to occur when divisions have  opened up within the political elite,  or between the  political elite and   the  bureaucratic layer:  in  such circumstances, spaces for alternative voices remained open, albeit constrained and subject to reversal.

Internationally, the  academic literature has  recognised the  fact  that ruling elites

have  expanded their repertories of social control beyond outright repression: as  a result, the  literature has  shifted away from focussing on  the concept of  repression, to that of pacification. According to Kein- scherf, pacification includes measures that “…produce undisruptive and unthreatening forms of collective action”.26  However, this is not to say that repression as it is commonly understood, and  pacification, are mutually exclusive: in fact, they  can  be complimentary strategies. For instance, the intelligence services can be used to separate out “good” protestors from  “bad”  protestors, and  the resulting protest  policing may  be  either facilitative or militarised depending on the type  of  risk  management strategies that the  police identify through the  intelligence gleaned.27  But the  fact  that the  elites  have found it necessary to shift from more visible to less visible  forms of social  containment at all, is not a sign of their strength; rather it is a sign of their weakness as they recognise the fact that they lack the capacity to repress openly. Why is this so? The next section will attempt to answer this question.

Organic Crisis: Growing Popular Capacity for Independent Action

It seems fair to say that South Africa is manifesting more elements of a classic Gramscian organic crisis.  For Gramsci, crises  become organic when they are thrown up directly by contradictions in how the capitalist system functions, when they  are  dynamic in that they  are  not  confined to particular actors, events, issues, or moments in time  or place, and  consequently when they  are a process rather than a  momentary eruption. The demands being raised may be diverse, and at times even incompatible. Such crises usually arise  when a particular regime of capitalist accumulation becomes unsustainable because of its own  internal contradictions. In such circumstances, the  ruling  bloc  (or the  coalition of interests that underpin a particular ruling  group) loses its legitimacy on a mass scale.  An organic crisis develops when the following conditions obtain:

Popular capacity for action increases;

More people can be detached from the previous hegemonic block and  be convinced to side with the subaltern classes;

There  is a decline in capacity of the elite to offer significant concessions, but;

There  is also a decline in the capacity of the hegemonic bloc to mobilise effective repression.

When  these conditions obtain, the  hegemonic  bloc  cannot offer  concessions easily, yet neither can it repress easily either.28

With  regard to  popular capacity for action  increasing, while  the  number of crowd management  incidents increased year-on-year since  1996,  too  little  can  be deduced from  this  upward swing,  as  the police database that logs  these incidents (the IRIS system) records both protests and gatherings. However, from  the  municipal data referred to earlier, it is apparent that protests peaked in  2011  (the  year  of the local government elections) in municipalities  such as eThekwini, Johannesburg and Lukhanji, which is when crowd management incidents recorded by SAPS peaked too.  So it is not  unreasonable to  assume that the peaking of incidents in 2011 can be attributed at least in part to an uptick in pro- test action, suggesting an increase in popular capacity for action as expressed through protests.

There  was little  evidence of co-ordination across protest sites, though; co-ordination occurred when a trade union movement organised a  national action, or  where a strike  took  place  in  different parts of the country, for instance a public sector strike. While  there was  little  evidence of  these protests coalescing into  more generalised political demands, they  have  the  potential to if a national political movement comes into  being that links  these different struggles together. The municipal data pointed to high  levels  of organisation, and  of new formations or even  organisations emerging all the  time,  suggesting that Patrick Bond’s term “popcorn protests”29 – used to describe seemingly sporadic, spontaneous protests ignored the  extent of organisation that actually exists.

There  was  no  evidence of unions and  community organisations uniting around shared grievances. However, it was apparent from  the municipal data that struggles at the  point of consumption are becoming as important to the  political life of direct action politics as struggles at the point of production, and  in some cases  (in the Makana municipality, for instance), the former are  overtaking the  latter as  flash- points of struggle.

When South African protests are viewed in  the  global  context, it becomes apparent  that popular capacity for action is not only  increasing, but  these increases are being sustained. The  protests could easily be  described as a cycle  in the  sense used by Tarrow.30

However, is the  protest cycle in  South Africa of even  greater historical significance? For instance, could it be part of a broader regional or even  global  protest wave? The question of whether the protests, including those in South Africa, are part of a wave,  rather than being isolated, single- country protest cycles, is an important one, as it speaks to whether the protests will fizzle out in time or escalate into fundamental and transformative challenges to the system on a worldwide scale. According to Colin Beck, the difference between a protest cycle and a protest wave is that the latter is present in at least  two or more societies within a decade of each other, and  these protests are tantamount to revolutionary situations.31 In other words, the protests affect more parts of the world over a longer period, and are not concentrated in a fixed period in time  or driven by a small,  well-defined set of actors. These features suggest that the unfolding struggles are responses to broader crises  in the world economy, and  in spite of their heterogeneity, they are capable of being sustained and even escalated into an insurrection precipitating an organic crisis.

Scores of Zimbabweans headed calls to peacefully demonstrate against the worsening economic woes Photo: Arthur Chatora

The mobilisations in Chiapas, the Occupy movement  in  the  United States, the  “pink  tide”  in Latin  America and  “Arab spring”, Palestinian struggles against Israeli occupation, and  anti-austerity protests in different parts of the  world,  are  all examples of challenges to the system in different regions of the world  (some more successful than others). Less well-known and  studied are the  wave of protests that engulfed sub- Saharan Africa in the  wake  of the Tunisian and  Egyptian political revolutions, with the most pronounced ones erupting in Swaziland, Mozambique, Zimbabwe and Malawi.

For Cox, if protest cycles  are sustained in more than one region of the world over a period of at least fifteen years, then this is a further indication that the crisis is organic, rather than episodic; as a result, the multiple resistances that have been mounted against the system could be described credibly as a revolutionary wave. Sustained regional disruptions usually happen at least once every twenty years.  The fact that some have  not led to regimes falling,  and  where political revolutions have  been achieved, they  have not  necessarily deepened into  social  revolutions, becomes less significant if revolutionary waves  are understood as a process rather than an event. If these protests have brought new political actors onto the streets, resulting in new forms of organisation, and extracted significant concessions from  ruling elites,  shaking the  state in the  process, then they  could be described as moving in an anti-systemic direction. This is because the  protests build confidence in the  power of collective action, and  consequently have the  potential to extract even  more significant concessions in future.32 When  viewed in this global  context, it becomes apparent that South Africa’s protests are  of world- historical significance, and  point towards them being part of a broader global  wave of heightened popular action. They are also likely to place  popular limits  on the  state’s ability  to use  organised violence, as doing so may well intensify popular action rather than dampen it.

In  fact,  while  the  protests cannot be said to have a distinct ideological character, the data points to the hegemony of the ANC diminishing.

With  respect to  more people detaching themselves from  the  hegemonic bloc, the  municipal data before 2011 pointed to the  ANC alliance dominating the  protest space – especially in smaller towns and rural areas – but  that its  dominance declined after  2011. The ANC alliance has proved to be a combustible one,  with  political alignments with  the  ruling  party coming under considerable pressure. At community levels, the municipal data suggested that the South African National Civic Organisation (Sanco) – often considered to be a fourth member of the ANC alliance – is largely a spent force, and  is being overtaken by a host  of independent community organisations or civics.  In  fact,  while  the  protests cannot be said to have a distinct ideological character, the data points to the hegemony of the ANC diminishing. This  does  not  support arguments advanced by Booysen and Fakir that the  protests are  merely about holding the ANC to account33: rather there is growing evidence of more communities becoming subjectively available for alternative politics to that offered by the ANC alliance.

Organic Crisis: Concessions or Repression

With respect to Gramsci’s two other conditions for an organic crisis,  namely that the elite  cannot offer concessions very easily, but  neither can they repress very easily, the neo-liberal phase of capitalism has entered a period of organic crisis in several regions of the  world.  This  phase is characterised by the financialisation of the economy, the rise  of  permanent mass unemployment and  declining rates of profit, creating conditions for a political crisis. In other words, these features make this  phase particularly unstable in that it creates conditions for mass revolt,  as fewer  concessions can be  offered than in  earlier expansionary periods (such as was the  case  under social democracy), while the  system cannot generate enough profit  to prevent itself  from contracting and  even  collapsing, worsening  the  socio-economic conditions even more.

In  the  case  of South Africa,  while the Zuma administration promised a more redistributive state, and undoubtedly many of its more principled office bearers remain subjectively committed to a more just and equal society, the  objective conditions in which they  came to office  did  not  favour radical redistribution.

South African President Jacob Zuma answers questions at the South African Parliament, on September 13, 2016, in Cape Town. Photo: ANP/ AFP Rodger Bosch

Yet at the  same time,  managers of the neo-liberal system – governments, financiers  and  other big  capitalists – need to maintain consent in  order to  continue ruling, which they  find  increasingly difficult  to obtain. If they  resort to coercion to stabilise the  system, they  risk  legitimacy and  state violence is used most effectively when consent remains for its use.34  Their inability to  resolve these crises  lie at  the heart of the  current period’s organic crisis. For instance, there are limits  on the  extent to which paramilitary policing can be used to  contain growing dissent. While  many police are  clearly  “getting away  with  murder’, public antipathy is building against the police and  the  political order they  seem to be propping up. There have not been nearly the  same levels  of protests against police violence in South Africa as there have  been US cities  such as Ferguson, in the  wake  of Michael Brown’s fatal shooting by the police. But, Marikana has hastened political shifts that have  been underway for  some time now,  and  has not  dampened protest levels: to that extent, it has not been a particularly successful massacre for the ruling  elite. The massacre was a precipitating factor in the formation of the Economic Freedom Fighters (EFF), and  Cosatu’s largest affiliate,  the National Union of Metalworkers of South Africa (Numsa), has been expelled from the federation, spurring it on to form  a United Front in  collaboration with  community organisations and  social  movements.

The state cannot risk going  even  further down this  path. Violence against dissenters on a mass scale is likely to eat into the ANC’s still considerable legitimacy, and hasten its slow but  steady demise at the polls.

Julius Malema, leader of the opposition party Economic Freedom Fighters (EFF), addresses supporters at the launch of the party’s local election manifesto in Soweto, near Johannesburg, South Africa, 30 April 2016 Photo: ANP/EPA Cornell Tukiri

This is especially so if a national movement comes into being that generalises protestors” demands, and  relates them to the  neo-liberal system of governance. Workers  and  the  poor, who face  the  brunt of the  system, are  increasingly unlikely to consent to supporting and funding their own oppression.

Another factor that  makes full-scale repression unlikely is that the security cluster  appears to  be  an  increasingly divided house, and   not   insignificant  cracks are beginning to  show.  The  police commissioner at the time of the Marikana massacre, Riah Phigeya, has been suspended and may well be dismissed for her role in the massacre.35  A spate of top  management resignations in the  SSA in 2011 has been linked to refusals to use the surveillance capacities of the  state to spy of Zuma’s detractors ahead of the  ANC’s elective conference in Mangaung.36  Furthermore, in order to  repress openly, the police would probably need the assistance of the  military. But the  military is industrialised and  unionised.

Another factor that makes full-scale repression unlikely is that the security cluster appears to be an increasingly divided house, and not insignificant cracks are beginning to show

In spite of arguments that unionisation can  compromise  combat-readiness, in 1999 the  South African  Constitutional Court legalised the formation of  military unions.37   There   is evidence that a significant number of soldiers  have  a consciousness of themselves, not  just  as  soldiers, but  as  workers, who are  exploited like other workers. Frustrations with  poor working conditions boiled over during the  ill-fated march on the  seat of political office,  the  Union Buildings, in August 2009. This very public confrontation – the  dynamics of which were misreported by many media organisations38 – led to the Chief of the Army, Solly Shoke, accusing the soldiers involved of mutiny, and  warning them that some other countries would have shot them for their actions.39

Former South African national Police Commissioner, Riah Phiyega. Photo: Werner Beukes/SAPA

In  view  of  these fractious relations, the  political elite  face a gamble: if the  current administration put  soldiers in  front of exploited, protesting workers (like the soldiers are),  and  told  them to shoot, what would they  do? What  if they  refused? Can they  really  risk a rebellion in the  military, which really would amount to mutiny? The political risks could be too great for them to gamble on the military.

The  more historically-aware security officials  are  also  likely  to  make political calculations about how  long  they  will last if they  intensify open repression. As the embarrassingly weak  presidency of Jacob Zuma splutters to  a close,  the  ignominious  fates  suffered by the  likes  of Saddam Hussein and   Muammar  Gaddafi, could well be top  of mind when they  attempt to weigh  up  the  long-term political costs  of engaging in an all-out defence of their positions. Regimes that relied on repression to maintain power have  never lasted.40  It is not coincidental that since  2009 – the  year that Zuma came to  office  – evidence has mounted of some even  resorting to political  assassinations to  silence their critics, especially in Mpumalanga and  KwaZulu/ Natal.41 The  2014 murder of Abahlali  base Mjondolo activist Thuli Ndlovu by two ANC councillors is a case in point.42 But resorting to informal repression is itself an indication that more violent sections of the  political elite  recognise that they  cannot engage in open repression. The  conviction of Ndlo- vu’s killers makes it less likely that political assassinations will escalate into  unstated state policy,  although as the country heads up to possibly its most fractious local government elections yet, it would be naïve  to ignore that the risks are there.

Conclusion

The  current government’s ability  to  offer meaningful concessions is limited, but  so too is its ability to repress easily. This means that the  South African  political landscape bears all the  hallmarks of having entered an  organic crisis.  Crises  of this  nature are not  necessarily negative; they  can  allow fundamental societal contradictions to surface in ways that force  society to confront them, grow from  them, and  move  forward. In  spite of fact  that the  current political moment seems so dark, the fact needs to be recognised that the  political space is wide open, and  is actually pregnant with  great promise. Does the state have the capacity to repress on a broader scale? It appears not. There  are unlikely to be more Marikanas in the  sense of an  organised, armed assault on protestors, although the possibility can- not  be ruled out  that state violence could occur as an unplanned reaction to particular events. While there are  clear  and  well- acknowledged legal limits  on  its ability  to use violence, the  political limits,  and  more specifically the  limits  imposed by popular agency, are less well-acknowledged.43

This is because repression is often studied as a static structural factor constraining movement activities, but not as a factor that is changed dynamically through interactions between state structures and  popular agency. Arguably,  the  social  and  political conditions that would allow the state to use ongoing (as opposed to sporadic) violence, do not  exist in this current conjuncture, as the  balance of power is shifting gradually towards popular movements outside the hegemonic bloc.  No matter how  powerful the men and  women with guns  seem, there are  important signs  that they  are  actually quite vulnerable, and  the shifting modes of repression point to that. While  overstatements about the  power of  the  coercive capacities of  the  state are  understand- able  in the  wake of Marikana, they  are not helpful, as they  can  lead  to fear,  and  even political paralysis. What Cox has referred to as “repression horror”44, can  lead  to movements seeing the  state as omniscient and omnipotent, even  when this  is, in fact, not the case.

But this does not mean that democratic movements must  become complacent. Appropriate activist strategies to  counter repression and  win back  democratic space are likely to be both timely and effective. On the other hand, ill-considered, misdirected ones, may  be  ineffective. In  this  regard, campaigns that  focus   on  the   account- ability  and  transparency of the intelligence services are  particularly important, as are campaigns to  defend the  right  to  protest from  administrative censorship (and not just  police violence). In  the  wake  of the Edward Snowden revelations, civil society and  social  movements are waging  a global fightback against unaccountable mass surveillance, and  already, they  have  won  significant victories. For instance, the  Barack Obama administration in the  US has  been forced to roll back  some of its most pernicious  mass surveillance programmes.45 But in South Africa, the  most effective method of limiting state violence in all its forms, is for movements to intensify popular organisation and  action, and  deepen the political shifts that are already underway.

This article is an edited  version of an inaugural  professorial lecture  delivered at  the  Council Chambers of  the  University  of  Johannesburg  on  13  July  2016.

This article was first published in Perspectives magazine by the Heinrich Boll Foundation