South Africa was hit hard by the pandemic. For a long time it ranked fifth in the world for total number of cases (behind countries with higher populations). The initial lockdown was draconian. The economy, already in recession, plummeted, and real unemployment rose from just under 40% of the workforce to about 50%.

Protest action in South Africa’s lockdown resembles a miniature of Geoffrey Pleyers’s sketch of the global battlefield [1]. As the lid was lifted, demonstrations and strikes began to surface, and a general strike is now planned. Here, some numbers add colour to his outline.

This country is noted for its high level of protests, as well as its massively high level of inequality, and here we detail what happened to the scale of action. Before that, however, we provide a brief note on how we reached our conclusions, and caveats about the findings.

Finding out what is happening

For 1997-2013 our analysis of protests in South Africa drew on data provided by the country’s police force (SAPS), which yielded a quantity and quality of information not matched anywhere in the world. For the period from 2012 statistics were derived from the Armed Conflict Location and Event Data project (ACLED).

ACLED is assiduous and rigorous, but it is overwhelmingly dependent on media reports, and these cover only a small minority of protests. Comparing our analyses of ACLED and SAPS data for the years 2012 and 2013, we found that SAPS protests out-numbered ACLED protests by more than 7 to 1.

There are several biases in media data. In particular, it is especially poor at capturing industrial protests, including strikes, and it is prone to ignoring orderly protests. Despite weaknesses, media data, specifically that gathered by ACLED, is valuable for assessing trends.

ACLED provides real-time data with notes on events and sources of evidence. Rather than rely on dashboard summaries, we worked with these notes, a painstaking task. For this, a protest was defined as ‘a popular mobilisation in support of a collective grievance.’ Rather than limiting ourselves to overall numbers of protests, we also investigated the grievances and the actors (i.e. those mobilized).

Turning to our findings, Graph 1 shows the frequency of protests from 2012 to 2019. Largely as a consequence of student mobilizations, there was a bulge in the years 2015 and 2016. The increase in 2019 is much steeper, reaching a total of 1473 protests recorded by ACLED. Of these, using our categorization of actors, 602 involved ‘residents’, up from 175 the year before. Most such protests were about service delivery.

Graph 2 shows the number of protests per week from the beginning of January 2020 until the end of July. The decline in March was dramatic, but that is partly because the peak in February was exceptional. There were 75 protests in January, quite a low number, but not uncommonly so - people were gradually returning after their annual break. Then there is a phenomenal rise in February. A short month. But 250 protests were recorded. This is more protests than in any month in 2019. We were heading towards a new peak in 2020, or so it seemed.

The number was still high in the first week of March. Then, news of COVID-19 spread, a state of national disaster was declared (15 March), and a lockdown was announced and then implemented (26 March). The number of protests came tumbling down, and ACLED did not record a single protest in the last 5 days of March.

In the end, there were still 92 protests in March, but only 33 in April. Since then there has been an upwards trend, with 74 in May, 144 in June, and 141 in July (a figure that might increase slightly with late recording).

Start with the actors

To interpret this pattern of events, we can start with actors, for which Graph 3 provides assistance. In our broader analysis we use 10 categories, applying them to countries across Africa. These are Residents, Workers, Students, Rights activists, Political, Religious/Cultural/Ethnic, Youth, Other, Multiple, and Unspecified. In South Africa, the actors seen most frequently in ACLED data come from the first three categories, and these are the ones depicted in the graph.

The massive rise in protests that occurred from the end of January to the beginning of March was largely a consequence of mobilisation by residents (community protests) and by students.

Among workers, throughout the period there was a trickle of disputes about work-place issues, often related to pay. There is a noticeable dip in April, the period of the hard lockdown (Level 5), when many workers were excluded from jobs and restrictions constrained organizational activity. These problems persisted to a degree, but there was a rise in the number of COVID-19 related disputes. The first of these were as early as 12 March, with demonstrations in two hospitals over shortage of personal protective equipment. Many similar disputes would follow. At the end of June there were also several protests involving drivers of minibus taxis, whose employers demanded state funds and/or the right to increase carrying capacity, and, in July, there were protests by public sector workers over pay, and by restaurant workers seeking to protect their livelihoods.

Early in the year, students raised a range of issues – the most common relating to administration of the National Student Financial Aid Scheme. But, on the weekend of 15 March, on-campus teaching was halted and residences were closed. Unsurprisingly, protests came to a sudden end.

Residents, students and unions

The majority of protests by residents were, as usual, about services. Lack of water was the predominant concern early in the year. Later on, with colder weather, disconnections and outages, electricity came to the fore. Criticism of councillors and demands over boundaries began. The second most common grouping of grievances was about crime – sometimes protests against policing, sometimes attacks on presumed criminals. There were many protests around gender-based violence, and this was the most commonly cited grievance in July.

There was also some action against ‘foreigners’, specifically in road haulage, but xenophobia also appears alongside other grievances.

Under lockdown, food became a major issue, with protests often taking the form of riots or attacks on trucks. There was also action around land and housing. Sometimes residents demanded re-location previously promised; sometimes, particularly recently, action was around forced ‘de-densification’; and sometimes residents mobilized against ‘land invasions’ by less-fortunate denizens. In July, there were also numerous protests calling for the closure of schools.

As is common in South Africa, protests often involved barricades of stones and burning tyres, and, barely less common, there was frequent use of rubber bullets by the police.

The analysis presented here suggests we are likely to witness a rising level of protests, at least by ‘residents’. Indeed, on 1 August, the C19 Peoples Coalition and its allies mobilized about 40 protests over various issues associated with the pandemic and lockdown. Individually the protests were quite small (with only one involving more than 200 participants), but they were probably more widespread geographically than any similar action since the end of apartheid.

Among students, while there is resentment about the way on-line teaching discriminates against poorer students, we are unlikely to see an upsurge in action during 2020. Among workers it is possible that increased unemployment will continue to sap confidence, but, as I write, on 28 August, there is a national strike by a large union active in the health and education sectors.

In the communities, however, the service delivery issues that fuelled the February protests remain. Further, poverty and unemployment have worsened, hunger has deepened and is more pervasive, and class divisions are even more stark. Deaths from COVID-19 are likely to be proportionately greater in poorer areas. There is extensive evidence of corruption, and the government is widely regarded as acting against the working class.

Already, as we can see in the statistics, there is growing unrest, and unless measures are taken to address the hardships people face, there is likely to be increased conflict and greater disorder. Indeed, the two largest union federations, the Congress of South African Trade Unions, aligned to the ruling African National Congress, and the left-wing South African Federation of Trade Unions have just called a general strike for 7 October.