INTERIM ELECTIONS REPORT Preliminary analysis of South African elections coverage for South African Broadcasting Corporation (SABC 2019) MEDIAMONITORING mum.? Saran Findlay and William Bird Media Monitoring Africa 2019 01 INTRODUCTION Elections are a critical moment in any democracy as they are a time in which citizens can hold their leaders accountable by either choosing to keep them in office or by deciding that other individuals or parties should take the reins. Within this, the media shoulder the fundamental responsibility of providing the electorate with all the information that they need to make informed choices about political party candidates so that citizens know how they can and should use their vote. This is even more important for a public broadcaster who is expected to push the boundaries in the type, range and depth of stories that they publish. In our local context, the South African Broadcasting Corporation (SABC) has been plagued by a myriad of political scandals, governance crises and persistent financial instability for the past 15 years. Given the extensive reach, control of the public broadcaster is seen as a fundamental stepping stone to accessing more power in the country and it has therefore been manipulated and used as a political tool time and again by those that seek to take over national leadership. 2016 Local Government Elections was a watershed moment and saw a slew of toxic decisions that affected the quality of journalism and safety ofjournalists at the broadcaster. While many important changes have taken place since then, we need to remain vigilant over the goings?on at the SABC. This interim report aims to give a brief overview of the type of coverage presented by SABC services during the 2019 elections period thus far. Another detailed report about the SABC's coverage will be released after election day. For more information, please feel free to contact us on (011) 788 - 1278 or at info@mma.org.za. 02 HOW WE CONDUCTED THE RESEARCH For this interim SABC report, we monitored all major news stations, including radio, television and SABC Online, across 10 of South Africa's official languages. Please see the table below for the full list of media monitored. For radio, we monitored three bulletins a day for each station (one in the morning 6am, one at 12 noon and one in the evening - 6pm). For television, we monitored one prime?time news bulletin in the evening. We also monitored SABC 2?s Morning Live and SABC 3's Democracy Gauge, as a full-length news programme and an elections?specific programme respectively. Within each bulletin or programme, we identified all stories related to the 2019 National and Provincial Elections in South Africa between 1 March and 14 April 2019. Information from each elections story was checked and recorded by specially trained monitors at Media Monitoring Africa (MMA) and the data was captured in specialized online software, Dexter. The data included: (1) name, type and origin of publication, (2) headline and summary, (3) main theme or topic of the story, (4) identity of sources (including name, race, gender and affiliation of individuals or groups who were accessed either directly or indirectly in the stories) and (5) whether any type of bias was present. The results below reflect the data from 983 elections items across the monitoring period. It must be noted that this is an interim report and a more extensive in? depth report about the elections coverage will be published after elections day. This should therefore not be taken a complete and flawless reflection of how the SABC has reported in this elections period, but merely as a chance to consider some initial findings. TABLE 1. FULL LIST OF MEDIA MONITORED RADIO TELEVISION ONLINE Lesedi fm SABC 1 Siswati/Ndebele News SABC Online News Motsweding SABC 1 lsizulu/Isixhosa News RSG SABC 2 Afrikaans News SAfm SABC 2 Sesotho/Setswana News Thobela fm SABC 2 Xitsonga/Tshivenda News Ukhozi fm SABC 3 English News Umhlobo Wenene fm SABC 3 Democracy Gauge SABC 2 Morning Live 03 03 WHAT WE FOU 0 WHAT WERE THE MAIN How are topics determined? In order tO gauge patterns in media coverage over time, we looked at the over? arching themes in stories across the period. For this, a primary topic was allocated tO each news item from an extensive pre?existing list. For example, the story ?Wednesday is D-Day for parties tO submit candidates lists" (SABC Onlfne, 12 March 20191) unpacks how political parties need to present their party candidate lists and pay the registration fee tO the Electoral Commission (IEC) in order tO participate on 8 May 2019. This story was primarily around the requirements and activities surrounding the elections, and was therefore identified as an "Elections logistics" item. In the same way, the story tO address its policies ahead Of 2019 national elections? (SABC Online, 18/03/20192) reports on the efforts undertaken by the PAC tO maintain party unity and to stop in?fighting ahead Of elections day. This item speaks primarily tO issues Of internal party politics and was therefore assigned ?Political party politics" as its key topic. ?gure 1. Top 5 topics across elections coverage in March 2019. .. .. NATIONAL POLITICS POLITICAL \m PARTY . PARTY POLITICS CAMPAIGNING A) NUMBER or SERVICE ELECTIONS ARTICLES DELIVERY ANALYSED CORRUPTION (15 O4 The graph (Figure 1) on the preceding page shows the breakdown of topics that received above 10% coverage across the SABC services. While some of the usual subjects, including political party politics and party campaigning, continue to receive extensive coverage, the graph is notable for two unusual additional topics. Corruption is an atypical topic but its prominence is largely due to the ongoing extensive coverage of the Zondo Commission, not necessarily due to a focus on parties dealing with corruption. Interestingly, and on a positive note, we see that service delivery is the other stand out topic. Here, a citizens? agenda appears to be shining through as issues of service delivery claimed the second highest spot at 15% of all coverage. This is an astonishing improvement compared to previous elections where service delivery made up only 1.75% and 0.6% of stories in 20163and 20144, respectively. Much of this reporting focused on the Eskom electricity crisis and the periodic spells of load?shedding experienced across the country as well as on the protests in Alexandra (Gauteng) where citizens took to the streets to engage politicians on issues of housing and refuse removal. Despite these being key contributors to coverage, other stories spotlighted different critical issues, including water, roads and housing. Encouragingly, these other stories were reported from across the country and gave a birds' view of various grievances, even from smaller communities and towns. This is particularly important for the mandate of the public broadcaster that seeks to offer South Africans insights into their own country. DELIVERY CLAIMED THE SECOND HIGHEST SPOT AT 15% OF ALL 05 WHOSE VOICES DO WE Recording and analysing the voices heard or sources provides us with insights into whose voices are amplified or indeed those who are silenced during an election period. Importantly, we identified sources as individuals who are both accessed (indirectly or indirectly) or who are simply mentioned in an item. A total of 2288 sources were recorded during the monitoring period resulting in an average ofjust over 2 sources per story Which groups are accessed in the coverage? Our findings show that despite diverse media being monitored, common groups of voices were repeatedly accessed across elections stories. Predictably, and following from the results in the previous section, political party sources alone made up over 40% of all sources. When combined with the Presidency as well as National and Local Government, political voices make up more than half of all voices heard across media. This means that politicians can easily set the elections agenda with their own narratives and ideas by contributing directly to the bulk of the views seen and heard by citizens. Having said this, citizens in this coverage comprised almost one-fifth of all sources and this is the highest level of access ever seen previouslys. Given how substantial this change is, it undoubtedly stems from remarkable efforts from internal decision?makers to shift the narrative away from the political elite to local people. While many of these citizen voices emanate from Democracy Gauge programming and may not totally reflect those voices in SABC news coverage, these findings still show how quality of coverage can be dramatically improved when the range of voices is deliberately amplified. ?gure 2. The most frequently accessed groups in elections coverage CITIZENS SA PRESIDENCY NATIONAL GOVT 19% 6% 5% POLITICAL PARTIES COREO-RATES 0 LOCAL GOVT 43 3% 2% 06 How did political parties fare? Media have the unenviable task of having to determine and provide fair, equitable coverage to political parties in an elections period. This is made all the more difficult given the explosion of new smaller parties contesting on 8 May as well as the public broadcaster?s mandate to afford equitable coverage to political parties. The top three political parties make up the lion's share of coverage with the ANC, DA and EFF totalling 70% of all political party sources. This is similar to patterns seen in 20144and 20163 elections. While these high levels of coverage are to be expected for the ruling party and the two main opposition parties, it is promising that the suite of other smaller parties were also given a voice in elections stories. At this stage, it is clear that the SABC services are affording the greatest diversity of political party voices than previously witnessed in coverage and the trend suggests strongly that the diversity is a direct result of efforts to ensure equitable coverage political parties. It will be interesting to see how coverage might shift, the closer to voting day we get. When coverage of the top three parties is compared to 2014, it is clear that the SABC is neither unduly favouring the top three parties through excessive coverage, nor is the SABC unduly disfavouring the top three through disproportionately less coverage. The spread of coverage is a positive shift from analysis in 2016 . where we found some services ?gure 3. Political party coverage clearly favouring the ANC and OTHER PARTIES clearly disfavouring the EFF and DA. These interim results for 2019 thus point to a clear effort by the SABC to afford UDM equitable and diverse coverage to the political parties participating in the 2019 CQPE elections. Clearly with 48 parties there will always be some level of dissatisfaction P?c with the coverage, especially amongst the smaller parties. Even then the SABC is to be IFP commended, thus far, for offering greater diversity than other media being monitored. 07 What was the race breakdown of voices? Given our country?s history of legalised race-based dispossession, analysing how different racial groups are accessed gives a clear indication of how much more work needs to be done to equalise the playing fields in terms of voice, access and power in South Africa. Comparing voices in the media to population estimates is a straightforward mechanism to do this. In this case, we can see that Black and White Sources are accessed largely in line with their size in South Africa?s population sitting at 80% and according to StatsSA6 respectively. Critically, though, Coloured voices only comprise half as much as they should be accessed (population: and this speaks to the ongoing marginalisation of Coloured people, their voices and their views. Repeated calls have been made across the country to recognise the rights of Coloured people. In contrast to this, Indian voices sit at double their population estimate of This over?representation stems largely from sourcing current Minister of Public Enterprises, Pravin Gordhan, and the recurring mentions of the Gupta brothers, Ajay, Tony and Rajesh, at the Zondo Commission of Inquiry into State Capture which is ongoing. Eigure 4. Breakdown of sources by race (o Eg} WW =?o=ao=so=io ?agWhite Coloured Indian voices voices v0Ices 08 What was the gender breakdown of those accessed? One of the key roles of any public broadcaster is to represent the diversity of its society. This section therefore aims to assess the SABC coverage and voices afforded by looking at the gender breakdown. Disappointingly, it appears that patriarchal biases that push for more men's voices in news remains largely intact in elections coverage. Here, men comprise an overwhelming 83% of all sources, despite making up less than half (249%) of those living in South Africa?. These findings are not unusual as men?s voices are favoured time and again over women's voices right across the globe7 . What is striking, however, is how dramatically under-accessed women are in SABC coverage compared to other South African media. recently published interim results of 2019 elections coverage revealed that 22% of sources were female8. It is critical that the public broadcaster, with the duty to lead in quality journalism, changes its approach to gender sources. In our subsequent report we will also provide a breakdown across the services. ?gure 5. Breakdown of all sources by_gender anme Male COMPRISE AN OVERWHELMING 83% OF ALL 09 0 HOW FAIR IS THE One of the most essential elements of ethical journalism is that of fairness and balance. In an elections period, this is made all the more important as biased reporting has the potential to undermine the freeness and fairness of the elections themselves. This is even more critical for South Africa?s public broadcaster who remains the most accessed source of news across the country. Given the recent history of the SABC where for the first time in 2016 MMA found clear evidence of systemic bias, the issue is of even greater importance. In monitoring, default position is one of fairness. i.e. the assumption is that all items are presented in a moderate, balanced and impartial manner until such a time as bias is identified. We specifically looked at bias in the form of language (exaggerated, generalised or trivialised), presentation or omission which sought to either highly favour or disfavour different parties. ?gure 6. Percentage of stories that Our interim results below were fair vs biased demonstrate the significant inroads taken to produce fair news and FAIR programming at the public broadcaster in 2019. In 2016, the SABC had shown horrifying results of systemic bias across some of its news platforms where only 90% of stories were fair9. While the socio? political climate at the SABC has changed fundamentally in many ways, this massive reduction BIAS in bias is still a hugely powerful statement in reclaiming the quality of the SABC and should be widely applauded. INTERIM RESULTS BELOW DEMONSTRATE SIGNIFICANT INROADS TAKEN TO PRODUCE FAIR 10 04 CONCLUSIONS This report clearly shows how much of the coverage maintains and perpetuates a political agenda both by widely accessed political voices as well as focusing on politicised subjects. Despite this, there are clearly some critical and deliberate efforts being made to include a greater diversity of voices, specifically citizens, as well as attempting to better represent South African racial groups. These signify institution-shifting inroads in the quality of SABC coverage that is unlike any other previous coverage. While some improvements are still needed, such as in the accessing of female sources, these interim results are highly encouraging for the SABC and for our impending elections. REFERENCES Article accessed here: candidates-lists Article accessed here: Media Monitoring Africa. 2016. Media Coverage of the 2016 Municipal Elections in South Africa. Accessiblehere: a/covera Medig Monitoring Africa. 2014. Media Coverage of the 2014 National and Provincial Elections in South Africa. Accessible Media Monitoring Africa. 2018. 15 years of reporting South African Elections Statistics SA, 2011. South African census. Global Media Monitoring Project (GMMP). Media Monitoring Africa. 2019. How are we doing? Preliminary analysis of South African elections coverage (March 2019). pp 13. Media Monitoring A rica. 2016. From protest ban to biased reporting? SABC Coverage of the 2016 Municipal Elections. 11