The ‘privatisation’ of municipal services has been a flash point for public policy debates around the world. From New York to Buenos Aires there have been loud and often heated disagreements about the merits and demerits of private sector involvement in essential services such as water and electricity. This is true of South Africa as well, where governments at all levels have been transforming the nature and scale of public ownership and public management.
Although the trend began in the late 1980s under the neo-apartheid National Party government, it was not until the election of the African National Congress (ANC) in 1994 – and the solidification of the ANC's electoral base with local government elections in 1995/6 – that the push to privatise services began in earnest. So thorough has this shift been that even the Free Market Foundation of South Africa can claim – with obvious delight – that the ‘privatisation of state-owned assets [has been] a guiding economic principle of South Africa's first democratically elected government’ (Spindler, 2004:1).
Strictly speaking, however, there has been relatively little outright privatisation in South Africa (and hence our use of quotation marks around the word). Although some state entities have been sold in whole or in part to private investors (e.g. Telkom) most private sector involvement in state services in South Africa has been in the form of public-private partnerships (PPPs).
With PPPs, the state typically owns the service assets (e.g. a water treatment plant) but the service may be managed and operated in whole or in part by a private company. Some PPPs involve small firms on short-term contracts (e.g. meter reading) while others may involve large multinationals responsible for all aspects of a service operation for more than 20 years. But regardless of size, PPPs have served to dramatically transform the philosophy and day-to-day operation of service delivery in South Africa. As a result, PPPs are often referred to as a form of privatisation in the critical literature on the subject because of the transfer of decision making and accountability from public to private hands (Barlow & Clarke, 2002; McDonald & Ruiters, 2005a; TNI, 2006).
Also of interest here are the creation of stand-alone ‘business units’ in South Africa – service entities that are wholly owned and operated by the state but which are setup at arm's length from municipal or national governments and expected to operate like a private company in terms of investment decisions and the recovery of operating costs (if not profits). This ‘corporatisation’ of services (strongest in water, waste management and electricity in South Africa) is not privatisation in the strict sense of the word either, but the results can be similar in terms of access to services for the poor and uneven patterns of investment and service quality (Shirley, 1999; McDonald & Pape, 2002; Smith, 2003).
In some circles in South Africa this privatisation and corporatisation of services has been met with much enthusiasm. The private sector itself has been supportive of these initiatives (with many large multinational service firms having established offices in the country) as have neo-liberal NGOs and research bodies (such as the Centre for Development and Enterprise). Funding agencies such as the World Bank and USAID have been active proponents as has the South African state which put considerable efforts into promoting privatisation and corporatisation via official websites, public information sessions, and advertisements in newspapers.
This support for ‘privatisation’ – in its various manifestations – forms part of a larger trend towards neo-liberal policy making in South Africa. While not uncontested in the country, there is a large and growing scholarship which argues that the South African state has become increasingly neo-liberal in its orientation and that this neo-liberal ideology has become increasingly hegemonic (Bond, 2000a, 2000b; Marais, 2000; Hart, 2002; Narsiah, 2002; Desai, 2003; McDonald, 2007).
But neo-liberalism in general, and privatisation in particular, has also met with much resistance in South Africa. The country's largest umbrella labour organisation – the Congress of South African Trade Unions (Cosatu) – has organised anti-privatisation strikes and marches involving several million workers. There has also been a mushrooming of community groups and non-governmental organisations opposed to privatisation (e.g. the Anti-Privatisation Forum, Anti-Eviction Campaign) as well as academics debating the matter.
Where has the South African press stood on the matter? In this article we look at the English-language print media over a four and a half year period, from May 2000 to October 2004 – a particularly intense time of debate over privatisation in the country and a period for which we had access to a comprehensive database of print media reporting. The focus of our study is the four core municipal services of water, electricity, sanitation and waste management, but the results have implications for privatisation initiatives in other sectors as well.
Our analysis involves a large database of newspaper articles, opinion pieces and letters-to-the-editor about these core services to determine the extent to which the print media could be deemed to be in favour of, or opposed to privatisation (as we have broadly defined it). We are also interested in the quality of this newspaper coverage (i.e. how informative and analytical it is), in part because of the confusion that sometimes appears in the press about what exactly constitutes privatisation. Also included is an examination of the more subtle, discursive aspects of this media coverage in order to comment on the potential impacts that the South African press may have had on popular perceptions around the privatisation of municipal services as well as on policy making. For this we develop a conceptual framework on hegemony, looking at the discursive tools employed by the media to construct and/ or strengthen dominant ideologies.
Our position is not a simple one. Although we argue that the oligopolistic corporate media in South Africa generates and perpetuates a neo-liberal discourse on privatisation, severely constraining the possibilities for alternative opinions to be heard or policies to be developed, this discourse is neither omnipotent nor monolithic. In fact, at first glance there would appear to be a relatively balanced coverage of privatisation debates in the South African English-language press, with an almost equal number of pro- and anti-privatisation articles in our sample, and many that attempt to provide ‘both sides’ of the argument. Alternative voices are heard and there are journalists and editors within the mainstream South African press that are sympathetic to anti-privatisation positions, if not in agreement.
But it is exactly this façade of balance, we argue, which gives neo-liberalism such hegemony in South Africa. By appearing to give equal space to different points of view there is a perception of ‘freedom of expression’ and ‘balanced reporting’ which obscures the more subtle, opinion-making discourses that generate biases in press coverage on this issue.
We begin our analysis with a brief conceptual overview of hegemony and the media in order to situate ourselves within this broad interpretive framework. This is followed by a description of our sample, the methodological techniques employed in our analysis, and their outcomes. We conclude with a brief discussion of how print media coverage may have affected popular and official discourse and policy making on privatisation in South Africa and what might be done to counter this neo-liberal hegemony.
Hegemony & Mass Media in South Africa
Marxist scholars have long argued that the media plays a key role in shaping public opinion in favour of the interests of capital (Althusser, 1971; Murdock & Golding, 1977; Hall, 1982; Murdock, 1982; Fairclough, 1989, 2001; Ferguson & Golding, 1997). Although Marx himself did not develop a detailed analysis of the processes and power relations that contribute to this dynamic, the following quote from The German Ideology (1968:21) is often cited in studies of mass communications:
The class which has the means of material production at its disposal has control at the same time over the means of mental production, so that thereby, generally speaking, the ideas of those who lack the means of mental production are subject to it.
This is not to suggest a deterministic view of the role that class plays in influencing the media. As Murdock and Golding (1977:19) note:
the process of ideological reproduction cannot be fully understood without an analysis of the economic context within which it takes place and of the pressures and determinations which this context exerts.
At one level we can look at the ownership structure of the media, controlled as it is in most countries today – South Africa included – by a handful of corporations with the same general interests as other fractions of capital in securing the conditions for rapid market growth through neo-liberal policy reforms such as liberalisation, privatisation and deregulation. In this respect, media operates like any other capitalist enterprise, seeking ways to improve its profit potential through expansion and cost-cutting and applying pressures on states for particular political and economic reforms (Murdock, 1982; Herman & McChesney, 1997).
The media is also a vehicle for capital in general in promoting itself (i.e. through advertising). Since advertising is typically the largest source of revenue for print and electronic media companies, it is in the interests of these media firms to create an ideological medium conducive to market-oriented growth. The latter can be done in part by ensuring market-friendly content within which advertisers would want to promote their products and services. Editors and journalists who share this market ideology are more likely to be hired and to have their material published, and the same can be said of opinion editorials (op-eds) and letters-to-the-editor – what Herman and Chomsky (1988) refer to as a form of ideological ‘gatekeeping’.
But on its own this is still an inadequate and overly deterministic account of the media and ideology. Most large media outlets operate in more subtle ways. As Murdock and Golding (1977:22) point out, the production of ideology is not purely economistic, arguing that ‘[Marx] used the notion of determination and conditioning not in the narrow sense but in a much looser sense of setting limits, exerting pressures and closing off options.’
Gramsci described this process as one of constructing rule by consent whereby capitalist classes, and those who benefit indirectly from their success, establish hegemony through the world of ideas rather than through (mere) force, creating a stronger and more stable form of rule. As Fairclough (2001:34) explains it:
There are … in gross terms two ways in which those who have power can exercise it and keep it: through coercing others to go along with them, with the ultimate sanctions of physical violence or death; or through winning others' consent to, or at least acquiescence in, their possession and exercise of power. In short, through coercion or consent. The state includes repressive forces which can be used to coerce if necessary, but any ruling class finds it less costly and less risky to rule if possible by consent. Ideology is the key mechanism of rule by consent, and because it is the favoured vehicle of ideology, discourse is of considerable social significance in this connection.
Peet (2001:57), in his analysis of neo-liberal hegemony in post-apartheid South Africa, points to the media as part of a cluster of institutions that ‘organise discursive flows’ and give ‘legitimacy’ to ideas. Other ‘centres of persuasion’ include universities, funding agencies, business federations, government departments, think tanks and private companies – what Peet refers to collectively as the ‘Academic-Institutional-Media (AIM) complex’. These institutions, according to Peet, have been ‘captured’ by neo-liberalism in South Africa, disseminating a consistent set of ideas shaped around a particular policy agenda. There are, argues Peet (2001:59), ‘competing tendencies within [this] AIM complex’,1 and counter hegemonic voices from outside it, but the end result is a David and Goliath battle of ideas, with research reports, conferences, policy papers, academic publications, donor funding and media coverage all being overwhelmingly neo-liberal in its orientation, serving to narrow (even crush) the potential for alternative ideas and discourses. The public impression of this debate, however, is one of objectivity, professionalism and a level playing field – a myth perpetuated by the dominant voices themselves. It is these complex, mediated and non-coercive spaces of public discourse that we want to unpack in this article.
Research Methods
Our discussion is based on a review of 647 news articles, op-eds and letters-to-the-editor taken from 34 English-language South African newspapers and wire services from May 2000 to October 2004 (see Appendix A for a list of all of the print media sources). The majority of these items were articles written by staff journalists (86% of the total), but there were a significant number of op-eds (9%), editorials (2%) and letters-to-the-editor (3%).
The sample was collated from an online database of print media collected by the Municipal Services Project (MSP) through a weekly search of websites of all major English-language newspapers and wire services (such as the South African Press Agency) as well as government websites which carry relevant press clippings (such as the Department of Water Affairs and Forestry).2 In some cases, this data was supplemented by articles from hard copies of major newspapers. All of the articles in the sample were explicitly related to the privatisation and corporatisation of municipal services in South Africa as they related to four core municipal services (water, electricity, sanitation and waste removal).
The sample provides a broad chronological representation of one of the most important and controversial periods of privatisation and commercialisation in South Africa. As Figure 1 demonstrates, the articles are spread throughout the review period, with spikes in coverage coinciding with pertinent events such as local government elections in December 2000, anti-privatisation strikes organised by Cosatu in August 2001 and October 2002, and various flare-ups in community resistance to service cutoffs and evictions in 2003 and 2004. The sample therefore provides good coverage of key public policy debates on the commercialisation of services.
There are, however, a few methodological limitations to our analysis. The first is that we confined our study to the text of articles, omitting photographs and other potentially important opinion-shaping representations. Second, millions of South Africans are functionally illiterate and may not, as a result, be affected by newspaper reportage,3 but the print media remains an influential source of news for policy makers in the country and impacts on other forms of media such as television and radio (Harber, 2002; Jacobs, 2004). Third, while English papers are the most dominant there are important isiZulu, Sesotho, Afrikaans and other Africanlanguage newspapers that help to shape public opinion. However, English is the lingua franca of political discourse in the country and with the advent of on-line newspapers the sphere of influence of the English-language press is only likely to grow in the future.
Pro- vs. Anti-Privatisation Articles
All 647 articles in our sample were reviewed and subjected to two interrelated content evaluations. The first was an assessment of each article's pro- or antiprivatisation position, ranked on a scale of -5 (strongly anti-privatisation) to +5 (strongly pro-privatisation). Pro-privatisation articles were characterised by statements, quotes, and/or information that promoted the agenda of corporatisation and/or privatisation. This included the promotion of all forms of private sector involvement in service delivery (e.g. outsourcing, management contracts, outright divestiture) as well as corporatised restructuring and the promotion of full cost recovery on services. Anti-privatisation articles, by contrast, were those that were largely critical of privatisation and corporatisation (to varying degrees). If an article gave equal/balanced coverage to both sides of the debate it was rated as neutral and given a score of 0.
In total, 41% of the sample was classified as pro-privatisation, 38% as antiprivatisation, and 21% as neutral. A breakdown by news agency showed some interesting variations, however, with Business Day printing by far the largest number of pro-privatisation pieces (102) and having the largest proportion of their coverage classified as pro-privatisation (62%) (see Figure 2 for data on news agencies with at least 10 articles in the sample). Other newspapers were more balanced in their coverage (e.g. the Sunday Times – the most widely read newspaper in the country – with an equal number of pro- and anti-privatisation articles). In a few cases the balance was tipped in favour of anti-privatisation coverage (e.g. the weekly Mail & Guardian where 54% of articles were assessed as anti-privatisation to some degree).
Analytical vs. Unanalytical Articles
Articles were then evaluated according to their degree of analysis (with -5 representing ‘very unanalytical’ and +5 representing ‘very analytical’, with gradations in between). An unanalytical article is one that would, for example, make a statement on the merits (or demerits) of privatisation without explaining what the author meant by ‘privatisation’ or how these arguments can play themselves out in practice. An analytical article, by contrast, would describe the particular form(s) of privatisation being discussed, the pros and cons of different forms of privatisation (or the similarities between them), and perhaps offer evidence to illustrate these points.
This supplementary analysis was added to our investigation exactly because so many writers did not clarify what they meant by privatisation or adequately explain the similarities/differences between its different variants. While recognising the difficulties of providing lengthy discussions of a complex subject matter, it was nonetheless possible to discern articles that made little or no effort to explain difficult or contentious topics and those that were able to provide some useful conceptual information and/or practical reference points.
In this supplementary analysis we begin to see a much less balanced picture of reportage, with more than two-thirds (69%) of the articles in our sample considered unanalytical and only 21% being considered analytical (with 10% being given a middle/neutral rating of 0). Figure 3 (over) provides an illustration of this assessment, sorted by news agencies with at least 10 articles in the sample.
The newspaper with the largest number of unanalytical pieces was Business Day, with 100 of its 164 articles (61%) falling into this category. The Mail & Guardian was the opposite with 46 of 68 pieces being characterised as analytical (63%) (the only newspaper in our sample to have a majority of its articles being classified this way).
This lack of analysis may be a reflection of a ‘dumbing down’ of journalism in South Africa – a phenomenon that Duncan (2003:8) blames on the low wages of journalists, but which has been raised in more general terms by others (de Beer & Steyn, 2002; Steyn & de Beer, 2004). It may also reflect an ingrained ideological bias amongst journalists and editors towards the marketisation of municipal services and a related willingness to allow unproblematised assumptions and positions on the topic to be published.
This latter point is demonstrated, in part, by the fact that there was a much larger proportion of pro-privatisation articles that were unanalytical than there were antiprivatisation articles. In other words, articles in favour of privatisation were much more likely to make unsubstantiated assumptions about the benefits of marketisation than anti-privatisation articles were to make unsubstantiated arguments about the negative aspects of marketisation. Pro-privatisation articles would, for example, speak to the ‘successes’ of a privatised venture but fail to mention any problems associated with it. Others would acknowledge past or potential problems but still argue that ‘there is no alternative’, leaving the reader with a very constrained understanding of the range of debates on the matter.
Anti-privatisation articles, op-eds and letters were not immune to this kind of unanalytical writing, sometimes spouting crude anti-privatisation rhetoric and offering little in the way of quantitative or qualitative scrutiny or comparative insight. For the most part, however, anti-privatisation pieces tended to be more substantive and reflective than their pro-privatisation counterparts. Not all of these ‘analytical’ pieces would be considered good journalism – there were often confused and problematic lines of argument – but the effort to provide context and analysis was much more evident and more consistent in the anti-privatisation cluster.
The higher quality of writing from those opposed to privatisation may be due to the fact that many of these articles were written by people not affiliated with a newspaper and who may have had the time and expertise to construct a more detailed, reflective article. The demands of the newsroom do not always allow for the degree of analysis one would like and we must be careful, therefore, not to be unfairly critical of busy, multi-tasking journalists and editors for not providing the same depth of analysis as a person writing a one-off letter or opinion piece.
This line of argument does not, however, explain the fact that many of the proprivatisation op-eds that we reviewed also lacked analytical rigour, falling into the same superficial and assumption-driven patterns as staff journalists. Not all proprivatisation opinion pieces and letters-to-the-editor fell into this category – some of the most sophisticated writing in our sample came from academics, government officials and think tanks in favour of privatisation – but the majority of proprivatisation articles fell into the unanalytical category, further illustrating the argument that ideas which fall within a dominant ideology may come in for less meticulous scrutiny in the English-language press.
Media Frames
We move now to the more discursive aspects of our review, beginning with an assessment of ‘media frames’, in an attempt to identify and assess the most dominant ideological concepts and arguments employed in the articles. Gamson and Modigliani (1989:3) define a media frame as ‘a central organizing idea, or frame, for making sense of relevant events, suggesting what is at issue’. Similarly, Entman (1993:52) describes frames as a way to ‘select some aspects of a perceived reality and make them more salient in a communicating text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation’.
Some media frames are intentional and explicitly stated, but often they are implicit, inadvertently capturing a particular world view. According to Gitlin (cited in Feree et al. 2002:14): ‘Frames, largely unspoken and unacknowledged, organise the world both for journalists who report it and, in some important degree, for us who rely on their reports’. When writers and editors select ideas and phrases to construct a story line, they are often buying in to a much wider world view which they may or may not believe in or understand themselves but which can convey powerful ideological messages.
We identified a total of 17 frames in our sample (nine of which fell into the proprivatisation camp and eight of which fell into the anti-privatisation camp), with a total of 1037 instances of these frames appearing in the sample (see Table 1 over). An initial list of 15 frames containing certain arguments for and against privatisation were identified after a preliminary reading of articles. As articles were analysed, any new frames that appeared were added to the list. The final list of 17 frames included two additional frames, ‘I’ and ‘L’ below. All frames within the articles fell into these 17 categories.
| Pro-privatisation Frames | |
| A | State-run services are inherently weak, unaccountable, corrupt, bankrupt and inefficient. There is ‘no alternative’ but to corporatise/privatise since the private sector brings efficiency and financial viability. |
| B | Privatisation is ‘pro-poor’ because it ensures better delivery for the needy by being more efficient, more accountable to the poor, and by providing the necessary capital for infrastructure investments and upgrades (capital that the public sector does not have). |
| C | Privatisation and corporatisation attract foreign investment, which helps the South African economy grow, thereby creating more jobs. |
| E | Making people pay for their services will help conserve scarce resources and help ‘consumers’ to see the ‘true’ economic value of the services they receive. |
| G | State-owned monopolies on service provision place risks in hand of taxpayers. It is better to shift that risk to the private sector. |
| H | People who criticise privatisation use incorrect information, manipulate the public, are trying to undermine the state/ANC and are out-of-step with the majority of South Africans. Included in this frame are anti-striking messages (e.g. that most workers are not behind the anti-privatisation strikes, that strikes are purely political, that unions refuse to engage meaningfully with government on these issues, and that they are undermining public services and the economy as a whole). |
| I | Government is privatising, but cares about and/or is working towards addressing the needs of South Africans living in poverty, including workers. |
| J | There are problems with privatisation, but they can be managed. |
| L | There is no need to worry because the government is not ‘privatising’; rather, it is establishing state-owned companies, facilitating public-private partnerships, outsourcing, and restructuring. |
| Anti-privatisation Frames | |
| N | Privatisation leads to corruption. |
| O | Communities are frustrated by privatisation-related policies and are resisting. There is a groundswell of resistance to privatisation and corporatisation. |
| P | Essential services such as water, sanitation, electricity, and waste removal are human rights and should be provided by the public sector with a focus on ensuring the service is provided rather than a focus on profit and cost recovery. Included in this frame is the argument that these services are Constitutional rights and that some privatisation and corporatisation initiatives are unconstitutional. |
| R | Under privatisation and corporatisation, service providers become less responsive to the needs of the poor because the focus shifts to cost recovery and profit, resulting in service cutoffs, limitations, evictions, and property seizure. Related to this frame are associated health impacts. |
| S | Under privatisation/corporatisation workers suffer through job losses and lower pay/benefits. Top executives, by contrast, see huge increases in salaries and benefits. |
| W | When services are outsourced or privatised the public has much less control over service provision as accountability shifts from elected officials to company management and shareholders. Included in this frame is the argument that resources such as water and electricity are public goods and should remain in public hands and be delivered by the public sector. |
| X | The implementation of aggressive cost recovery under corporatisation and privatisation has exacerbated the extreme inequities inherited from the apartheid system (the rich get richer and the poor get poorer). |
| Y | International financial institutions, governments, and multinational corporations pressure governments all over the world through loans, aid, and debt relief to privatize and attract foreign investment. Included in this frame is the argument that privatisation facilitates an economic relationship that is another form of colonialism. |
All of the articles in our sample had at least one clearly discernable frame, and many had two (hence the total of 1037). Those with more than one frame tended to be either pro- or anti-privatisation, but some articles had pro-and anti-privatisation frames (another indication, perhaps, of the somewhat confused and largely unanalytical nature of journalism on this topic in South Africa). There were 522 instances of proprivatisation frames and 515 instances of anti-privatisation frames in the sample.
On the pro-privatisation side, the most frequent frame was ‘A’ (‘there is no alternative to privatisation’). At 220 occurrences it dominated the pro-privatisation discourse and was the most frequent frame of the entire sample. Frame A is also the most ideologically loaded and inflexible of the frame concepts. The argument, essentially, is that the public sector is inherently corrupt and inefficient and that there is no point in exploring alternatives that do not involve the private sector or private sector operating principles.
It is difficult to say with certainty what impact this line of argument might have on South African newspaper readers, but given the frequency with which this frame appeared, and the ‘neutrality’ that might be ascribed to it (given that in most instances this frame was delivered by regular journalists and not by opinion writers), it could be a major contributor to a public perception that privatisation is ‘common sense’.
The other three most common pro-privatisation frames are also revealing. Frame B (‘privatisation is ‘pro-poor’ because it leads to better service delivery’) appeared 73 different times, and fits with the ANC's (and the World Bank's) position that commercialisation is being done in the interests of the poor. This is a direct counter attack to the anti-privatisation argument that privatisation is being done in the interests of capital and elites, and has come to dominate official government discourse on all service delivery reforms – at all levels of government – and has crept into the neo-liberal academic literature as well (see for example SACN, 2004). In some respects this is a sign that the dominant discourse has felt pressure from competing discourses, changing its language to deflect anti-poor criticisms. Fairclough (1989) calls this ‘creative restructuring’ of language.
Frame C (‘privatisation attracts foreign investment and is good for the economy and jobs’) occurred 58 times in the sample and can be seen as another attempt to directly counter anti-privatisation critics who have argued that private capital is not interested in job creation and that privatisation results in net job losses. Here we see a commitment to the official government line that commercialisation is good for the South African economy, combined with a faith in the ability of the private sector to produce jobs.
Frame H (‘people who criticise privatisation use incorrect information, manipulate the public and are out-of-step with the majority of South Africans’) appeared 61 times. With this frame we see an attempt to discredit the intellectual skills and personal integrity of anti-privatisation critics and to make these groups and individuals appear anti-ANC; even unpatriotic. While not an entirely unexpected or unique debating tactic for a controversial public policy matter such as this, the use of this frame was frequently mean-spirited in its tone (even from ‘objective’ staff journalists), making for remarkably heated attacks on anti-privatisation critics while contributing little to the substantive questions at hand. The targets of these attacks included union officials, community-based organisations, academics and NGOs, with ‘ultra-leftist’ being one of the most popular ways to attempt to discredit a person or organisation (other terms include ‘infantile leftists’, ‘self-styled leaders’ and ‘short-minded individuals’).4
Some writers even portrayed anti-privatisation groups and individuals as immoral and unlawful and as intentionally misrepresenting facts. Much of the source of this language was from government officials, with journalists unquestioningly reproducing highly controversial statements of a personal/ideological nature without questioning their validity or providing context. One example is the following remarks made by the former Minister of Public Enterprises, Jeff Radebe, in 2001, widely reproduced verbatim in the press:
The so-called Soweto Electricity Crisis Committee has become part of this criminal culture [of challenging privatisation by not paying for services]. Its campaign has resulted in the destruction of Eskom's assets [the parastatal electricity provider] … The representatives of [this group] have proven that they will do anything, including telling lies to the community, in order to realise its political ends. Such people cannot be regarded as the genuine representatives of our people.5
Anti-privatisation articles were not immune from these kinds of ad hominemattacks, but they tended, on the whole, to be more focused on substantive issues. The most common anti-privatisation frame was R (‘the private sector is less responsive to the needs of low-income people’), with 216 occurrences. This was followed by frame S (‘workers suffer through job losses, lower pay and fewer benefits’) which occurred 162 times, and P (‘essential services are human/Constitutional rights and should be provided by the public sector’) which occurred 35 times. These anti-privatisation frames were most common in op-eds and letters-to-the-editor, raising the question once again as to how they would be interpreted by readers and whether they would be seen as less ‘objective’ than articles by staff journalists (though there were sophisticated anti-privatisation frames employed by staff journalists as well).
It is also useful to note how these frames appeared chronologically in relation to key privatisation events (such as the Cosatu-organised anti-privatisation strikes in August 2001 and October 2002). Figure 4 illustrates a timeline for the most common pro- and anti-privatisation frames (A, B, R and S). From this timeline it is evident that particular media frames were employed around specific events, with Frame S appearing to be the most incident-sensitive (another indication of the opinion-based character of much of the anti-privatisation writing). The pro-privatisation frames peaked around these events as well, but frame A (‘there is no alternative’) appeared consistently throughout the period under study, with the most sustained impact on newspaper readers and public opinion.
Critical Language Study
Our final look at the data provides the most abstract, but potentially most important, insight into media coverage of privatisation. Here we look at the actual terminology employed in the media frames to further illustrate the deep ideological bias that permeates pro-privatisation journalism on the subject.
We have adopted a method of discourse analysis outlined by Fairclough (1989), in which he argues that language is used strategically – though not always intentionally – to compete for ideological position, and as a tool to exert power over others. For Fairclough (1989:32), this is a struggle between classes that manifests itself – in part – as discourse, with mass media being a key terrain: ‘The way in which orders of discourse are structured, and the ideologies which they embody, are determined by relationships of power in particular social institutions, and in society as a whole. We therefore need to be sensitive in critical language analysis to properties of the society and institutions we are concerned with’. The objective of this analysis is to examine words and grammatical choices for their expressive, experiential, and relational values. Fairclough (1989:112) describes these different values as follows: Experiential – ‘a trace of and a cue to the way in which the text producer's experience of the natural or social world is represented’; Relational – ‘a trace of and a cue to the social relationships which are enacted via the text in the discourse’; and Expressive – values dealing with ‘subjects and social identities’.
When analysing a text's vocabulary for experiential values, one looks for the ‘classification schemes’ that reflect a particular discourse (e.g., privatisation, efficiency, financial viability, cost recovery are all part of classification schemes common to neo-liberalism). In addition, ‘rewording’, ‘overwording’, ‘ideologically contested words’, and ‘ideologically significant meaning relations’ are also possible clues to the way discourses manifest within a text (Fairclough, 1989:110-113).
When analysing a text's vocabulary for relational value, analysts look for euphemisms and any ‘markedly formal or informal words’. In examining a vocabulary for expressive values, particular attention should be paid to what the author might be trying to express by choosing one particular word instead of other word options. It is important here to look for metaphors and their ideological significance (Fairclough, 1989:116-120).
Grammar must also be scrutinised for experiential, relational, and expressive values. Fairclough suggests that we identify which types of process and participants predominate, whether or not agency is clear, and whether or not sentences are active or passive. For example, an article might say, ‘The municipality cut off water to residents of this township,’ or it might say, ‘Water was cut off to residents of this township.’ In the latter, agency is unclear, leaving the reader without an explicit understanding of who was responsible for the cutoffs.
We applied all of these methodological tools in a detailed analysis of a sub-sample of our data set (30% of the articles), helping us uncover subtle (and not so subtle) attempts to present privatisation initiatives in their best possible light, to make public sector options appear inherently flawed, and to villanise opponents of privatisation.
To illustrate these points we have selected a few examples from our sample. The first comes from a Financial Mail article in November 2000 (‘Water Privatisation Incentives Go Under the Bridge in Johannesburg’), written by a staff journalist in what appears to have been an attempt to write a balanced review of the decision by Johannesburg City Council to contract out the management of its water services to a private consortium. The article was rated as +1 (pro-privatisation) and -2 (unanalytical) in our coding system.
In the article the journalist reports that, ‘The council's advisers told Johannesburg that if the water utility were run efficiently, revenue over the five-year life of the contract could range between R350m and R1.4bn’. The positive attributes of the term ‘efficiency’ are used here in relation to the contract and illustrate elements of a neoliberal classification scheme, implying that the city will save enormous sums of money by using a private management firm, while at the same time suggesting that the previous (public) system was inefficient.
The article also states that the private partners have ‘committed to turning around the Water & Sanitation Utility of Johannesburg … and transfer[ing] skills to a new, black management team’. Here it is argued that the private consortium is ‘committed’ to upliftment, particularly for black South Africans, making the contract a pro-poor and socially progressive development (unlike the previous public sector arrangement, presumably). The journalist goes on to note that the corporation seeking the contract ‘took account of local conditions’ and ‘has experience in countries similar to SA’, demonstrating its ‘commitment’ to socio-economic performance and making it a ‘charitable’, ‘world-class’ operator. All of this phrasing is an implicit or explicit endorsement of this particular privatisation initiative, despite the journalist's attempt to provide a neutral account of the contract.
A pro-privatisation article from The Mercury (‘Durban Hit by 28% Rise in Price of Water’, 29 May 2001) illustrates a different set of discursive trends – ones which attempt to blame the poor for weak public sector performance and to criminalise their activities. In this piece the journalist notes that, ‘The city's water department, sitting with a R96 million accumulated deficit, has incurred further costs running into millions’, and goes on to blame these losses on ‘townships which used water without paying for the service’, quoting the executive director of Durban metro water services as saying that ‘non-payment for water by some residents had forced the council to operate on a bank overdraft’. The implication here is that township residents are free-riders and a drain on the municipal coffers, with no effort being made by the journalist to situate ‘non-payment’ within the larger political and economic context of struggles over water delivery and other basic services in Durban. The notion of a ‘culture of non-payment’ is used to blame the poor for public sector inefficiencies, despite the fact that research has increasingly demonstrated that it is ‘an inability to pay, rather than an unwillingness to pay, that lies behind poor payment compliance’ (SACN, 2004:91; see also McDonald, 2002).
But perhaps the single most important language problem relates to the use of the word ‘privatisation’ itself. Seldom employed in pro-privatisation articles, one finds instead much softer euphemisms such as ‘partnerships’ and ‘restructuring’ (although the institutional make-up of these arrangements are often left unexplained). The same applies to ‘corporatisation’, which is typically referred to as a ‘business unit’. Stripped of inherent political or economic meaning, and imbued with neutral, even positive, connotations (who could be opposed to a ‘partnership’?) this new privatisation discourse diverts attention away from the deeply ‘private’ and/or ‘corporate’ character of many of these transformative efforts and the negative inferences often associated with these words (for an extended discussion of why the term ‘privatisation’ remains an appropriate way to define different forms of commercialisation see McDonald & Ruiters (2005b:14-24)).
To be fair, journalists and editors have not invented this new language – much of it can be traced to the World Bank and the South African government – but the press has picked up on this new terminology and uncritically used it or reproduced it in their coverage of privatisation debates, serving to de-politicise (or, more accurately, re-politicise) the very meaning and intent of commercialisation efforts.
Even more striking, perhaps, is that elements of this neo-liberal discourse appear in many of the anti-privatisation articles as well. In some cases this would appear to be the result of journalists or op-ed writers reporting on, or expressing an opinion on, anti-privatisation positions but offering, in the end, a confused and sometimes contradictory set of arguments and terminology to express them.
Use of the terms ‘partnership’ and ‘efficiency’ feature dominantly here, but many of the other neo-liberal words and phrases described above were found in antiprivatisation writing. One example is the use of language that blames and/or criminalises the poor. In an otherwise strongly anti-privatisation article in the Daily News (‘Chatsworth Tenants Declare War on City Authorities’, 21 May 2001), the author refers to township residents ‘flouting the law’ and ‘illegally reconnecting electricity supplies’. Rather than portraying community protests against commercialisation as acts of civil disobedience, and ones aimed at obtaining what many see to be Constitutional rights to basic services at an affordable cost (Leibenberg, 2004; Flynn & Chirwa, 2005), anti-privatisation protestors are characterised as ‘angry tenants’.
In some cases this use of neo-liberal language by anti-privatisation writers was intentional. Typically this was the product of a journalist or editor who claimed to be opposed to privatisation but insisted that public-private partnerships, corporatisation and other forms of commercialisation do not constitute ‘privatisation’ per se. The editor of the Mail & Guardian made just such a claim in a feature article in June 2004, arguing that it is ‘plain wrong’ to say that ‘South Africa has undergone a massive exercise in water and electricity privatisation’. Instead, she argues, ‘[o]nly four of 284 municipalities – and relatively small ones at that – have contracted out the management of water. As for electricity supply – none, none of it has been privatised’. The writer fails to mention that some five million people are serviced by these privatised water systems in South Africa and that virtually every water and electricity utility in the country has been corporatised to some extent.
Conclusion
In conclusion, the English-language press in South Africa has provided what would appear, on the surface, to be a relatively balanced coverage of the privatisation debates over core municipal services, with an almost equal number of articles for and against privatisation and many that attempt to provide ‘both sides’ of the argument. But the quality of this news coverage, the source of the news items, the kinds of information packages used to explain the debate, and the actual phrases and terminologies employed, have the effect of solidifying privatisation as a dominant discourse, turning the tide in its favour and limiting the possibilities of anti-privatisation discourse and debate.
The degree to which this media coverage influences public opinion and policy making is difficult to ascertain with certainty, but it is an important part of a larger neo-liberal shift that has altered the political and ideological landscape of South Africa over the past 15 years and which has helped to pave the way for the privatisation and commercialisation of core municipal services in the country.
But having made this argument about the importance of text in policy debates we close the paper with an important caveat, cautioning readers against an overemphasis on discourse analysis and highlighting the need for continued investigation into the more material aspects of privatisation. As Harvey (1996:80) notes in his review of discursive research:
To privilege discourse above other moments is insufficient, misleading, and even dangerous. Errors arise when examination of one ‘moment’ is held sufficient to understand the totality of the social process. Again and again we will find slippages of the sort that convert a dialectically correct statement like ‘there is nothing outside of the text’ into false statements that ‘everything can be understood through texts’ (or, worse still, ‘everything is a text and can be understood as such’) and equally false practices that seek to use, say, the deconstruction of texts as the privileged (and sometimes the only) pathway to understanding.
It is essential, therefore, not to lose sight of the fact that there is a larger material basis for privatisation in South Africa. Understanding the political economy of what Harvey (2003) calls ‘accumulation by dispossession’ must remain at the heart of privatisation investigations and challenges. In this respect it is essential to see privatisation as part of a larger accumulation strategy in South Africa; one that assists capital with access to large and relatively untapped markets for capital investment and profit, providing an outlet for overaccumulated domestic and foreign capital, creating opportunities for multi-tiered service delivery and helping to create conditions for a wider mass consumption society (Bond, 2000a; McDonald, 2007).
The link we want to stress here is the connection between text and the reproduction of these material relations. Language is not, in and of itself, a driver of privatisation, but it plays an integral (and dialectical) part in shaping the ways in which capital is able to penetrate the service sector and legitimate its role therein. It is essential to make the links between these larger (and ever changing) structural features of privatisation and their discursive manifestations.