Deliberative Democracy and Electoral Reform in South Africa: A Campus Experiment

ABSTRACT

In the transition to an inclusive democracy, South Africa changed its electoral system for the national parliament to proportional representation. Ever since, there have been suggestions of electoral reform. So far, the debate has rarely involved ordinary citizens. This article presents the results of a Deliberation Day on Electoral Reform in South Africa at the University of Cape Town. This campus experiment in deliberative democracy was part of a project-based course for MA students in Political Studies. The outcomes are as predicted in the literature: the knowledge of the 47 student participants increased and their opinions became more coherent. The student participants were highly critical about South Africa’s political system and demanded more accountability through the electoral system. The most striking, and encouraging, outcome was the unanimous support for more deliberation on campus. The findings presented here suggest the potential of deliberative democracy for organising the national debate on electoral reform as well as for communication inside South Africa’s universities.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1 “Deliberative poll” is a registered trademark of prof. Fishkin at Stanford University.

2 See: http://cdd.stanford.edu.

3 See: http://www.cmu.edu/student-affairs/dean/acad_int/conversations.

4 At the time, I was the Van Zyl Slabbert Visiting Professor in the Department of Political Studies at the UCT. I would not have been able to teach the course and do the project without the crucial support of key people, first of all prof. Anthony Butler, the head of the Department of Political Studies, and prof. Robert Mattes, the director of the Centre for Social Science Research. Mrs Gajjar and Mrs Polzin provided invaluable administrative support. Matthias Krönke assisted with data entry. The MA students in my course helped prepare and organise Deliberation Day from beginning to end and were instrumental to its success as moderators of the small-group sessions: Darkowa Awinador-Kanyirige, Christopher Edyegu, Kristýna Greplová, Richard Griffin, Zahira Grimwood, Leonard Mbulle Nziege, and Davide Rasconi. Previous versions of this article were presented in the Africa Research Group at the Central European University and the 2016 General Conference of the European Consortium for Political Research (ECPR) in Prague. I thank the participants, as well as André Bächtiger, for helpful comments.

5 This requirement, which fits with the general preference for proportional representation after violent conflict (Bogaards 2013), has limited the options for electoral reform (cf. Bogaards 2007). Non-proportional electoral systems are ruled out from the start and among proportional electoral systems, those that rely on preferential voting are considered too complicated.

6 There are many more criteria for electoral system design. For an overview, see Reynolds, Reilly, and Ellis (2005).

7 More than 4 out of 10 participants in Deliberation Day admitted not to know the name of their constituency representative. The same number provided guesses that were obviously incorrect.

8 Technically, even now, South Africa's electoral system is two-tier PR (Elklit and Roberts 1996).

9 Already before the first inclusive elections, Reynolds (1993) advocated a similar system, with slightly larger districts (5–12 members).

10 See the Minutes of the Discussion of the Electoral Amendment Bill in the Home Affairs Committee of the Parliament of South Africa. Available at: https://pmg.org.za/committee-meeting/16044.

11 See: “Electoral Amendment Bill.” http://www.gov.za/sites/www.gov.za/files/36267_gen233.pdf.”

12 For example, Hamilton (2014) favours a German-style electoral system for South Africa. His mention of Ireland might be mistaken for a recommendation of the single-transferable vote practiced in both parts of the island, but from the description, it is clear that this is based on a misunderstanding and that Hamilton endorses a German-style combination of PR and elections in individual constituencies.

13 Likewise, Cooper-Knock (2016) sees institutional reform as only part of the solution to South Africa's many problems and Zuern (2015) expects the most from social contention. 

14 Following UCT protocol, the Department of Student Affairs approved this study, after a positive review of the Ethics Committee of the Department of Political Studies. UCT policies regarding access to student records did not allow for the random sampling that is normally recommended (Fishkin 2009).

15 Of the 21 students who completed the pre-deliberation survey but did not attend Deliberation Day, 19 gave permission to use their responses. A comparison of participants and non-participants shows that although they are similar in background features, they differ in many other respects. Non-participants are more negative about the state of democracy in South Africa and the working of the electoral system. They are also less committed to the ideal of democracy. They score worse on knowledge questions and also report feeling less politically competent. Remarkably, 60% are not sure what kind of electoral system they prefer, against 15% among the participants to Deliberation Day. On the other hand, they have no problem answering more specific questions about the electoral system and they show the same preferences as participants (e.g. against an electoral threshold, in favour of independent candidates). In sum, the students who completed the pre-deliberation survey but did not attend Deliberation Day are overall more critical and insecure than the participants. Arguably, it is precisely these students that could have benefitted most from the deliberative experience.

16 49 students participated in Deliberation Day and completed the post-deliberation survey. For two of the participants we could not identify a matching pre-deliberation survey and they are excluded from the analysis.

17 The exception is identification with the Economic Freedom Fighters (EFF), which is in any case more popular with young people. According to the sixth round of the Afrobarometer, with data for 2014/2015, 12% of respondents aged 15–29 felt close to the EFF. Available at http://www.afrobarometer.org.

18 For an empirical analysis of the impact on small-group deliberation of race and gender composition, see, respectively, Sommers (2006) and Karpowitz, Mendelberg, and Shaker (2012).

19 For problems with self-selection, see the documentation of the various Campus Conversations conducted at Carnegie Mellon University. Melville, Willingham, and Dedrick (2005, 53) admit that the National Issues Forums in the USA ‘often do not represent a true cross section of the community’. At the University of Limpopo, Oyedemi and Mahlatji (2016) conducted a qualitative study on voter turnout among born frees with a focus group of 18 self-selected students.

20 The preliminary results of the Deliberation Day on Electoral Reform in South Africa were presented at a roundtable in mid-December 2015 with academics, including two former consultants of the ETT; a leading activist in the non-governmental organisation My Vote Counts; and the convenor of the UCT Global Citizenship Program. To enhance transparency, all student participants and experts were invited to this event.

21 Overall, young people are more likely to change their mind, demonstrating the kind of open-mindedness that deliberation requires (Suiter, Farrell, and O’Malley 2014).

22 This figure comes from the sixth round of the Afrobarometer (2014/2015) in South Africa, available at http://www.afrobarometer.org. See also Gouws and Schulz-Herzenberg (2016, 23–25).

23 In the expectation that not all participants might have had a chance to read the brochure, we included a ‘cannot really say’ answer. Three out of the 47 participants chose this option.

24 Available at: http://caae.phil.cmu.edu/cc/CC_Handbook_Final.pdf.

25 The six moderators, two for each of the three deliberating groups, were all students in my course. They were prepared for their task by a professional from the Institute for Justice and Reconciliation in Cape Town who also chaired the Q&A session during Deliberation Day.

26 See Mendelberg (2002) and Landwehr and Holzinger (2010) on the difference that types of outcome make for the process of deliberation and Baccaro et al. (2016) on the impact of discussion modalities.

27 Deliberative experiments of this kind take place in a highly controlled environment and one might wonder whether this control could not be used for political ends. For example, moderators ‘can be seen as potentially having the power to control the process and to guide the deliberation in a specific direction’ (Elstub 2014, 176). However, there is no indication that the participants felt manipulated to arrive at a particular outcome. Several survey questions on the quality of deliberation directly asked about the fairness of the background information, the impartiality of the moderators, and the helpfulness of the experts. No problems were flagged.

28 The reference numbers come from the sixth round of the Afrobarometer (2014/2015) in South Africa, available at http://www.afrobarometer.org.

Additional information

Funding

This work was supported by the Van Zyl Slabbert visiting chair at UCT.