The Evolution of the Middle Class During Nearly Two Decades of Democracy in South Africa

ABSTRACT

South Africa entered the Third Wave of democracy in 1994. Since then it has successfully faced many challenges, the last one being the global financial crisis of 2008–2009, without much damage to its democratic institutions. Pessimism about its democratic trajectory is largely the result of unfortunate policy choices resulting in very low economic growth. Among the many factors that assist democracies to stabilise and endure is the role of the middle class: the more extensive the middle class, the greater the chance for democratic stability. This should be reflected in their values and attitudes toward democracy and institutions. In this article, we use survey data from 1995 and 2013 in order to analyse the evolution of the middle class in South Africa and its values and attitudes since the first democratic elections. Results indicate that the South African middle class has doubled in size during these two decades and that the internal racial distribution has grown closer to the ethnic distribution of the country. We nevertheless conclude that in South Africa democracy is not yet ‘the only game in town’.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 Bhorat explains that the Gini coefficient is ‘the measure of income inequality, ranging from 0 to 1. Zero (0) is a perfectly equal society and a value of 1 represents a perfectly unequal society’. According to Statistics SA, the

Gini coefficient (income inequality) has declined from 0,72% in 2006 to 0,68% in 2015; however, there are notable variations amongst various population groups. Black Africans have the highest income inequality with a Gini coefficient of 0,65 in 2015, increasing from 0,64 in 2006. Income inequality amongst whites declined from 0,56 in 2006 to 0,51 in 2015. The Gini coefficient amongst Coloureds declined from 0,60 in 2006 to 0,58 in 2015. Despite having experienced declines in income inequality in 2009 (0,53) and 2011 (0,50), the Gini coefficient for Indian/Asians was 0,56 for 2006 and 2015. (Statistics SA 2017)

2 See also Ikenberry (2013) for an update on the issue. By regions, see Tridico (2013) on Central and Eastern European countries; Fernandes and Heller (2006) on India; and Kaabi (2012) on Turkey and Saudi Arabia.

3 On the meaning of middle class in areas like Soweto see, for instance, Alexander et al. (2013) or Phadi and Ceruti (2011). Basically, the label middle class seems to “denote self-sufficiency, responsibility and social mobility” (Phadi and Ceruti 2011, 88).

4 This term has acquired negative connotations over time and became derogatory. Originally, ‘it was coined by TNS Research Surveys (Pty) Ltd and the UCT Unilever Institute to refer to members of South Africa's fast-growing, affluent and influential black community’. See https://en.wikipedia.org/wiki/Black_Diamonds_(racial_term) .

5 On the middle class in Kenya see, for instance, Burbidge (2014). On this issue in other regions or countries such as Thailand, see Baker (2016); on China, Nathan (2016); on India, Lahiri (2015); on Chile, Barrueto and Navia (2015).

6 See García-Rivero, Kotzé, and du Toit (2003, 9–12) for a detailed discussion on this. See also Hertova, López-Calva, and Ortiz-Juarez (2010, 5) and Southall (2014, 522).

7 The Masters and Servants Acts of 1856 had been passed in 1856 already, which made the following criminal offences: breach of contract breach, desertion or striking. This was basically applied to unskilled workers, who were mostly Black people. In 1911 the Mines and Works Act No. 12 was passed, reserving certain skilled jobs for Whites and Coloureds; the Industrial Conciliation Act of 1924 provided for job reservation; the Mines and Works Act (Colour Bar Act) No. 25 of 1926 provided for certificates of competency for skilled work, and the Natives Labour (Settlement of Disputes) Act of 1953 prohibited strikes for Blacks, to mention but a few of the many pieces of legislation to keep the non-white labour force out of skilled jobs.

8 On measuring class through occupation, see Goldthorpe and Erikson (1993), Evans and Mills (1999) or Hadjar and Samuel (2015) to mention but a few.

9 Middle range income refers to those with an income located between the two and four income quintiles. Income item from surveys has been divided into five quintiles, being the second and fourth the upper and lower limit to consider those self-employed whose income falls within that range as ‘old’ middle class. This is done following the method applied by García-Rivero, Kotzé, and du Toit (2003); see also Rubinson and Quilan (1977, 615).

10 For technical details, see http://www.worldvaluessurvey.org/WVSDocumentationWV6.jsp. In 2013, the sample size was N = 3531 and representative of the adult population with a statistical margin of error of less than 2% (at 95% confidence level). The 1995 data were from the World Values Survey with an adult sample size of N = 2935.

11 The question was:

Here is a scale of incomes. We would like to know in which group your household is, counting all wages, salaries, pensions and other incomes that come in per month. Just give us the letter of the group hour household falls into, before taxes and other deductions.

12 The Indian group shows the second highest income mean (12.000–13.900 Rand) after the White population (16.000–17.999 Rand).

13 LSM was computed by weighting 29 items people have in their households, e.g. hot water, flush toilet, refrigerator, vacuum cleaner, TV, air conditioner, etc. and divides the population into 10 groups. For this analysis, it has been recoded into four groups, namely, lowest class – unemployed, low middle class, middle class and upper class. Detailed information about LSM can be found at http://www.saarf.co.za/lsm/lsms.asp. The World Values Survey includes this item in the questionnaire.

14 The question read: ‘People sometimes describe themselves as belonging to the working class, the middle class, or the upper or lower class. Would you describe yourself as belonging to the … ?’ Possible responses: Upper class; Upper middle class; Lower middle class; Working class; Lower class; Don't know/Refused.

15 In a group break-down, there are some similarities and discrepancies. Probably with a larger database, more precise results would be achieved.

16 Basically, ‘[i]nstitutional trust [ìs] a crucial dimension in the link between political culture and democracy, There is thus ample consensus that institutional confidence is necessary for political stability in established democracies as well as the consolidation of new democracies’ (Kotzè and Garcia Rivero 2017, 2). On the importance of confidence in institutions for democracy, see also Gibson (2001), Inglehart (1990), Lagos (1997) and Rice and Feldman (1997).

17 In both 1995 and 2013 surveys, the item reads: ‘Having a democratic political system’. Values: 1 ‘Very bad’ – 4 ‘Very good’.

18 The question reads: How much confidence do you have in the following institutions? Response options ranged from ‘quite a lot’ to ‘not at all’ – five categories.