The impact of subjective social status, inequality perceptions, and inequality tolerance on demand for redistribution. The case of a highly unequal society

Abstract

The international literature shows that demand for redistribution is influenced by subjective factors like perceptions of inequality and individuals’ perceived social status. However, few have assessed these subjective dynamics in the developing South, especially in Africa. This study aims to assess the subjective interlinkages between subjective social status, inequality perceptions and demand for redistribution in South Africa, the country with the highest level of inequality in the world. Using ISSP data and an ordered probit model, we show that subjective social status is a negative and significant determinant of demand for redistribution in South Africa. This means that South Africans who position themselves on the lower rungs of society demand higher redistribution and vice versa. Furthermore, perceptions of inequality are significant in driving demand for redistribution in South Africa, as this study finds that inequality tolerance negatively influences demand for redistribution and perceptions of actual inequality positively influence demand for redistribution. This means that South Africans who perceive higher levels of inequality tend to demand more redistribution, while South Africans who tolerate more inequality tend to demand less redistribution. Surprisingly, inequality tolerance is relatively high among the unemployed, Africans, and females. Many of these individuals are part of the most vulnerable in society and would actually benefit from more redistribution. Overall, the results show that, in South Africa, subjective factors like subjective social status and attitudes towards inequality significantly influence demand for redistribution.

Keywords:

• Social stratification
• social norms
• inequality
• redistribution

JEL CLASSIFICATION CODES:

• D3
• H3
• Z13

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 The 2001 population census enumerator areas (EAs) were used as PSUs.

2 It should be noted that there are various measures for perceived inequality. In his paper on perceived inequality, Roberts (2014) used the Differences in incomes are too large question in the ISSP survey to measure perceived inequality. Although we use a different measure of perceived inequality, the measure used by Roberts is still incorporated as a robustness check for the impact of perceived inequality on demand for redistribution.

3 On the top are the wage estimates for doctors in general practice, a cabinet minister in the national government and a chairman of a large national corporation, while at the bottom are the wage estimations for an unskilled factory worker and a shop assistant.

4 Based on the ordinal nature of the dependent variable, marginal effects are only reported for the base outcome of 5, meaning those who have the highest demand for redistribution.

5 To perform the various parallel-line assumption tests, the ordered logit model needs to be estimated. Since the coefficients of the ordered logit and probit are very similar, the study also performs an ordered logit to perform the post estimation parallel-line assumption tests that include the Brant, Wald, Likelihood-ratio score and Wolfe Gould tests. These ordered logit regression results can be found in the Appendix.

6 Analysis of Variance (ANOVA) was used to test whether the mean demand for redistribution of the population groups were different and statistically significant. The results confirm statistically significant differences for all population groups and results can be found in the Table A2 in the Appendix.