Are South Africa's Protests Really Driven by Rising Expectations?

By Jane Duncan · 6 Oct 2014

A+ A= A-
    Print this page       comments
Picture: Defence Web
Picture: Defence Web

Are South Africans protesting because government service delivery is poor? Or are they protesting because delivery is so good that expectations have been raised to the point where government cannot meet them?

The ‘rising expectation’ explanation of the protests has found favour with a diverse range of institutions and individuals, such as the government, the South African Institute for Race Relations (SAIRR), City Press editor Ferial Haffajee and Municipal IQ. It featured prominently in President Jacob Zuma’s most recent State of the Nation address.

This argument is fascinating, as it allows vested interests to perform a rhetorical sleight-of-hand, turning a ‘bad news story’ into a ‘good news story’. ‘Rising expectations’ allows government to argue that protests occur, not because the government is doing badly on service delivery, but because it is doing well.

The SAIRR also have a vested interest in spinning the protests, as they support more neoliberal solutions to South Africa’s problems than the ANC government. To this end, the Institute has argued that the provision of free services creates expectations that all services should be free. The only way for the government to free itself from this trap, it argues, is to privatise service provision to reduce unrealistic and unsustainable demands on the public purse.

An institution which measures the performance of municipalities, Municipal IQ, has also climbed on the rising expectation bandwagon, although they have also been critical about its explanatory value, too. It has also given credence to the argument’s ugly twin, ‘relative deprivation theory’, which maintains that social movements are inspired to struggle because they see services being delivered in other areas, but not in their own.

In a recent City Press article on the protests, Municipal IQ said that they found it strange that there were more protests in provinces where service delivery was most pronounced, notably Gauteng and the Western Cape. Conversely, they argued, protests levels were low in provinces where service delivery was practically non-existent, such as the Eastern Cape and KwaZulu-Natal.  Yet the evidence provided by City Press’s own feature - which analysed service delivery in protest hotspots - was inconclusive on the link between protests and service delivery.

Police statistics problematise the Institute’s claims. According to statistics released to Media24 Investigations, between 2009 and 2012, Gauteng saw the largest number of service delivery protests, followed by the North West province, KwaZulu/ Natal and Mpumalanga. The Northern Cape had the lowest number of protests, followed by the Western Cape. While it is unsurprising that Gauteng, as South Africa’s industrial heartland, registers the largest number of protests, the police figures for the other provinces do not bear Municipal IQ’s arguments out.

The province with the second highest number of protests (according to the police), the North-West, can hardly be described as an epicentre of service delivery; in fact, corruption and mismanagement appears to be distributed fairly uniformly across the province. Yet, if the ‘rising expectation’ argument was to be taken seriously, the Western Cape would be brimming with protests.

What explains the discrepancy between Municipal IQ’s statistics and the police’s? The police compile their own statistics from public order police records, and these are then pooled in a database called the Incident Registration Information System.

Municipal IQ has explained that their protest estimates are lower than those of the police as they collect data about protests against municipalities only. In doing so, they make the problematic assumption that only municipalities are responsible for service delivery. Yet the methodological problems do not stop there.

According to a Municipal IQ press release, the organisation relies on media reports for its protest database. If this is the case, then it stands to reason that it would record more protests in the Western Cape and Gauteng, as these are the most media-rich provinces. Other provinces may be less visible in Municipal IQ’s protest space, not because there are fewer protests, but because the paucity of media may lead to protests being under-reported. This methodological deficiency should caution journalists against taking Municipal IQ’s arguments, and the statistics on which they are based, at face value.

Undoubtedly, the government’s record of delivering services, in the sense of installing services, is admirable, and Statistics South Africa’s data bears this out. But the rising expectation argument assumes that mere availability of services is enough to measure the universality of service delivery: so, accessibility and affordability don’t really matter. Yet so many protests have been about these delivery dimensions, and ‘rising expectation’ proponents remain blind to these realities.

This narrow approach towards measuring universality doesn’t take into account the fact that some provincial and local governments have failed to maintain the infrastructure that has been rolled out, leading to disruptions in service delivery. The commodification of service delivery in the early 2000’s led to water, electricity and telecommunications services becoming increasingly unaffordable, and the free basic services on offer were grossly inadequate.

‘Rising expectations’ also assumes that the protests are overwhelmingly about service delivery, when this is not the case. This explanation flattens out the diversity of protests, and caricatures them in unhelpful ways. Other issues sparking protests include crime, industrial disputes, housing and education.

Many protestors also feel aggrieved about being inappropriately represented, or even misrepresented. The University of Johannesburg’s South African Research Chair in Social Change has also linked migration in the North-West to the province’s propensity to protest. Mono-causal explanations, such as ‘rising expectations’, fail to capture these complexities.

Case studies of protests also unsettle the ‘rising expectation’ claim as a general explanation. According to the Social Change Chair’s research, the 2009 protests in Balfour suggested that protestors wanted to relocate from Mpumalanga to Gauteng because they aspired to having the level of services offered in the apparently better-off Heidelberg. So there was evidence supporting the relative deprivation argument that delivery elsewhere fuelled rising expectations in Balfour. Yet relative deprivation does not explain why, in the face of uneven development, some communities engage in collective action and not others.

Three years after the Balfour protests, destructive protests broke out in Heidelberg, debunking the view that the town was a model for service delivery. In the process of objecting to electricity disconnections and the cost of basic services, protestors burnt down houses, municipal buildings, and an old age home. Rising expectations cannot explain these protests, as according to this argument, Heidelberg residents had no reason to protest in the first place.

There are dangers in allowing public debates about the protests to be dominated by conservative vested interests. One danger is that the protests themselves are likely to intensify and become more destructive. This is because protestors may feel that their grievances are not being heard or are being distorted, which may increase desperation.

Furthermore, explaining protests largely in terms that make the government look good can create space for official explanations that posit a sinister ‘third force’ as protest drivers. This is especially so in municipalities that have a good delivery track record, narrowly defined. After all, the argument goes, these are not ‘real’ protests motivated by ‘real’ grievances. Needless to say, such explanations serve the more securitised elements of the ANC government well, as they can be used to delegitimise the protests and justify repressive responses.

Protests in South Africa are big news at the moment, so it is important to ‘get the story right’. Lives may even depend on it. In this regard, journalists should treat the rising expectation argument with caution.

While claiming to be a general explanation of the protests, this argument is expedient and self-serving, and when held up against the reality of protests, it actually has unclear, even limited, explanatory value. And that should be the bottom line for any journalist seeking expert analysis to interpret a complex social phenomenon.

** Correction: This article was corrected on 9 October 2014. It had previously stated "In a recent City Press article on the protests, Municipal IQ said that they found it strange that there were more protests in provinces where service delivery was most pronounced, notably Gauteng and the Eastern Cape. Conversely, they argued, protests levels were low in provinces where service delivery was practically non-existent, such as the Eastern Cape and the Free State." This statement has been amended in the article.

Duncan is a Professor of Journalism at the University of Johannesburg.

Should you wish to republish this SACSIS article, please attribute the author and cite The South African Civil Society Information Service as its source.

All of SACSIS' originally produced articles, videos, podcasts and transcripts are licensed under a Creative Commons license. For more information about our Copyright Policy, please click here.

To receive an email notification when a new SACSIS article is published, please click here.

For regular and timely updates of new SACSIS articles, you can also follow us on Twitter @SACSIS_News and/or become a SACSIS fan on Facebook.

You can find this page online at

A+ A= A-
    Print this page       comments

Leave A Comment

Posts by unregistered readers are moderated. Posts by registered readers are published immediately. Why wait? Register now or log in!


6 Oct

The Rising Expectations Argument

Perhaps the protests are not about service delivery although that is how the protesters may explain their behaviour. It could be something real but which somehow remain unsaid. Poverty, hunger and unemployment are some of the possible real causes of the endless and seemingly puzzling protests in SA.

Respond to this comment

Barry Saayman
6 Oct

Honest Expression of Opinion, without Malice or Dishonest Motives

The Press Code states-

7.3 Comment by the press shall be an honest expression of opinion, without malice or dishonest motives, and shall take fair account of all available facts which are material to the matter commented upon.

9. Violence - Due care and responsibility shall be exercised by the press with regard to the presentation of brutality, violence and suffering.

I wonder if subservient news outlets supporting their chosen loyalties, including ideologies, can claim that they are without malice or dishonest motives and that they in fact exercise due care and responsibility.

>>"In a recent City Press article on the protests, Municipal IQ said that they found it strange that there were more protests in provinces where service delivery was most pronounced, notably Gauteng and the Eastern Cape..."

Must probably read Western Cape and not Eastern Cape.

>>"And that should be the bottom line for any journalist seeking expert analysis to interpret a complex social phenomenon."

Spinning is the questionable strategy of some journalists, certain "experts" and Non-Governmental Organisations fronting for either the government of the day or big business.

What is particularly tormenting is the observation that these "experts" normally spin as far as they can believe themselves and believing one's own propaganda represents in my view the pinnacle of stupidity. Who else but themselves do they fool all the time?

SA will never solve any of the serious problems that our broken and long suffering society face, including rampant violence, if problem areas and/or the causes for recurring disturbing phenomena are misdiagnosed, denied or "successfully" rationalised away by "experts" and spinned into "good stories" by politicians, their associates, leading civil servants and civil society and supporting media.

Repeating the denials and rationalisations of "friends" with little or no credibility does not represent balanced reporting.

Both argumentum ad verecundiam and argumentum ad populum are logical fallacies and professional journalists should be acutely aware of this.

Somehow some of those that should have been defenders and friends of an open society, namely journalists and lawyers, have morphed into its fiercest enemies in South Africa and this state of affairs already have serious consequences. The freedom that individualists love dearly is under siege.

"The media's the most powerful entity on earth. They have the power to make the innocent guilty and to make the guilty innocent, and that's power. Because they control the minds of the masses." - Malcolm X

A well written column by Prof Duncan

Respond to this comment

Municipal IQ
9 Oct

Response to Jane Duncan (Part One)

What causes service delivery protests? A response to factual inaccuracies by Municipal IQ to Jane Duncan's article "Are South Africa's protests really driven by rising expectations?" (Part One)

Municipal IQ has compiled data on service delivery protests staged against local government since its launch in 2007. We have over ten years of data based on publically available sources, predominantly media reports of protests. Despite the title of the article, most of the content is focussed on attacking Municipal IQ's methodology and questioning our credibility as a source of data and analysis on service delivery protests.

For this reason we find it necessary to respond to a number of arguments posed by Duncan in her article.

1. Is relative deprivation a reasonable explanation of service delivery protests?

Duncan questions whether one group feeling relatively more deprived than a neighbouring group may be a cause of service delivery protests. While there can be various interpretations as to why protests take place in well-performing municipalities, and even wards, we find it credible that one needs to understand the city-level context in which one community feels less well-served than a neighbour. Last month for example, a Nelson Mandela Bay community staged a protest with the express grievance that they saw a contractor commence work on a housing project in neighbouring Winnie Madikizela Village. Duncan also seems unaware that our view on relative deprivation (rather than absolute deprivation) is based on extensive quantitative socio-economic analysis against municipal, provincial and national averages.

2. Is SAPS 'crowd-related' data an accurate measure of protest activity at a local level?

Duncan uses the South African Police Service (SAPS) 'crowd-related' incidence data as an alternative source of data to question the accuracy of Municipal IQ service delivery data. However, it is clear that Duncan misunderstand the SAPS data. SAPS has never stated that its 'crowd-related' incidence data is a measure of service delivery protests although it is sometimes erroneously quoted in the media as such. Our own analysis of the SAPS 'crowd-related' incidence data shows that while this data might INCLUDE service delivery protests, it also includes ALL OTHER crowd control incidents in a locality, including issues as diverse as monitoring political meetings in halls, marches protesting rising crime in an area, sports even and so forth; in fact this data records any incident involving a group of people in a locality where SAPS had a presence. This also explains why in 2013 the SAPS data recorded 12 399 'crowd-related' incidents, while Municipal IQ only recorded 155 service delivery protests. Duncan wouldn't be the first to conflate SAPS data on 'crowd-related' incidents with service delivery protests, but a clearer analysis of this issue would perhaps help her to appreciate why the North West, site of Marikana conflict and much industrial unrest (but not municipal protests), would tally up a markedly different proportional representation in SAPS crowd-incidence data than in Municipal IQ's results.

Respond to this comment

Jane Duncan
12 Oct

Response to Municipal IQ's response

In response to 1, this still raises the question of how Municipal IQ quantifies the protests.

In response to 2, Municipal IQ is incorrect in stating that I relied on SAPS crowd-related incidence data an alternative source to question the accuracy of Municipal IQ service delivery data. If Municipal IQ followed the link provided in my original article, they would discover that SAPS themselves provided service delivery protest data only to Media 24.

Furthermore, if Municipal IQ consulted the following article (, they would realise that I too have criticised the conflation of the SAPS crowd control data with protest data, and would not make the mistake they accuse me of making.

Municipal IQ
9 Oct

Response to Jane Duncan (Part Two)

What causes service delivery protests? A response to factual inaccuracies by Municipal IQ to Jane Duncan's article "Are South Africa's protests really driven by rising expectations?" (Part Two)

3. Is a monitor built on media reports an inadequate measure of protest activity?

Municipal IQ relies on media reports and other sources for its protest monitor and would be the first to acknowledge that there is a risk that as Duncan states, the "paucity of media may lead to protests being under-reported [in less well covered provinces]". Does possible under-reporting render our research valueless as Duncan implies? Our results and trends are confirmed by other sources of data on service delivery protests including other organisations monitoring protests such as the University of the Western Cape's Multi Level Government Initiative.

We have never suggested that we have definitive insight into protests, but rather that we use public domain sources to provide a proxy for protest trends to add to and enrich debates around service delivery protests. Our analysis has been able to support discussions on whether seasonality is a factor in protests from one year to another, in which provinces protests have been seen to escalate or recede and whether periods before, during or after elections are prone to protests, using the same, consistent methodology over a ten-year period.

4. Does the 'rising expectation' theory also support 'third force' explanations of protest activity?

Duncan argues that the "rising expectation" theory also supports "third force" explanations of protest activity. While it seems contradictory that in Duncan's world of sweeping generalisations one can at the same time believe protests are caused by rising expectations as well as a third force, Municipal IQ has in numerous newspaper articles come out against the notion of a third force causing protests, especially given the evidence from our protests data that shows a wide variation of protest dynamics, their often random nature and diversity.

5. Are protests about service delivery or something else?

Duncan also claims that the "'Rising expectations' [explanation] also assumes that the protests are overwhelmingly about service delivery, when this is not the case". What evidence does Duncan provide for the claim that protests are not caused by service delivery? Interestingly, her own research into "records of notifications sent by organisations intending to hold gatherings and protests" in 7 municipalities ( Leaving aside Duncan's assumption that 'notice of protests' is an adequate measure of protest activity (highly questionable given that most protests take place without permission), to claim with such confidence that a study of protest causes in 7 municipalities can be used to generalise protest causes in all 278 municipalities, would require an astounding leap of faith.

Municipal IQ finds it ironic that in the same article that she urges journalists "to get the story right", Duncan fails dismally to do this herself, misquoting, quoting selectively, misinterpreting information and inappropriately providing her own research to overstate her view.

Perhaps instead of focusing most of her attention on Municipal IQ, Duncan should rather have tried to unpack the complex question she initially posed.

Respond to this comment

Jane Duncan
12 Oct

Response to Municipal IQ

Hmm, now the comment in 3 is interesting. Municipal IQ states here that that they rely on media reports "and other sources". This contradicts their own press statement which I linked to, as well as this article: Both make it explicitly clear that they rely on media articles. I find it interesting and revealing that the phrase "and other sources" has suddenly crept into their vocabulary. Media reports are very valuable for qualitative research, but they are not to be relied on for quantitative research, including quantifying the number of protests, and especially the number that have turned violent. I explain the reasons why here:

On 4., Municipal IQ has noted the limits of the rising expectation argument in the article I link to. On balance, though, their publicly available pronouncements make my statement that they have also climbed on the rising expectation bandwagon, although they have also been critical about its explanatory value, fair comment. But they cannot deny that their research has been used to prop up pro-government arguments about how wonderful service delivery is (See here:, which creates space for these arguments to be made. This article links Municipal IQ both to rising expectation and relative deprivation arguments.

Another article that's more categorical in its endorsement of the rising expectation argument, written by Municipal IQ, says the following:

'Urbanisation, essentially the influx of poor migrants to cities, is prompted by a search for jobs, and therefore is most pronounced in areas of economic growth. But this results in an irony - although service delivery protests are commonly perceived as an indication of a failure of local government, Municipal IQ has found a strong link between municipal productivity (a measure of local government success) and service delivery protests - those in search of jobs move to successful cities where they perceive there to be economic opportunity'. Apart from the Jeremy Cronin article linked to above, this has been re-quoted here

If they have misrepresented Municipal IQ's views about rising expectations being a credible general explanatory argument, notwithstanding the caveats that they themselves have mentioned in the article that I linked to, then I would imagine that they demanded corrections and/or retractions from these organisations.

On 5., Municipal IQ is correct to say that notices to municipalities may not take account of the protests that took place outside the ambit of the Regulation of Gatherings Act, but if they had read further, they would have realised that this is acknowledged. In her notes to the article I link to, the M&G journalist who wrote the article says the following: "The municipal data analysed includes information about protests that were cancelled or were denied permission but it does not tell us which protests turned violent or which went ahead without notifying the authorities in accordance with the Act". Again, see the first article in the series ( It notes the need to triangulate data sets to overcome these methodological weaknesses.

Records of applications from 7 municipalities out of 278 are enough to problematise the arguments that protests are overwhelmingly about service delivery, and it doesn't at all require an astounding leap of faith when set against other research that has taken place on the same subject and when considering the methodological difficulties of acquiring the municipal data. Accessing this data set is incredibly time consuming, costly and tedious, as it involves going to municipalities and manually inputting data from all notifications.

In the light of this problem, the data set is going to be small, but it is a reliable quantitative data set available about the reasons for protests as it comes from the "horses' mouth" so to speak, in that its culled from the notifications written by the conveners.

The SA Research Chair in Social Change has also found a diversity in causes of protests in their own independent research. SAPS's own data as given to Media24 shows that service delivery protests constitute a fraction of crowd control incidents they monitor, bearing in mind the caveat that crowd control incidents include gatherings and protests.