Copyright © 2022
Inclusive Society Institute
50 Long Street
Cape Town, 8000
South Africa
235-515 NPO
All rights reserved. No part of this publication may be reproduced or transmitted in any form
or by any means without the permission in writing from the Inclusive Society Institute.
DISCLAIMER
Views expressed in this report do not necessarily represent the views of the
Inclusive Society Institute or those of their respective Board or Council members.
Authors: Percept and 71point4
Editor: Daryl Swanepoel
7 February 2022
Executive summary
“What does inequality, poverty and socio-demographic change look like in South Africa – at present, and over the past 27 years of democracy?”
This report aims to address that question. To do so, we take a multidimensional approach to assessing inequality (as well as poverty and socio-demographic change), considering inequality in eight dimensions, namely: Economic, Education, Health, Living Conditions, Social and Cultural, Physical Security and Legal, Political, and Environmental.
For each of the dimensions, we review the literature and analyse data from various sources to provide a view at present, as well as over time.
Our approach moves one beyond referring to a Gini coefficient metric to answer the question of inequality. A shift towards a multidimensional framing of poverty and inequality in South Africa lends itself to a more comprehensive description of what has, and has not, been achieved during the new democratic dispensation. By interrogating the changes made in South Africa in a more granular way, the lived experience of South Africans may be better understood and reflected on. The approach also allows the drivers of inequality to be identified. In this way, levers of change and opportunity can be described, and poverty can be tackled more comprehensively.
Where useful, we rely on frameworks to consider how all of the mentioned dimensions link together. Some dimensions relate to ‘fundamental capabilities’ (like equal access to health, education, and economic participation). Others relate to ‘functionings’ (like health and education outcomes, and income and wealth). And yet others relate to ‘conversion factors’ (for example, personal, socio-political and environmental circumstances), where the latter may influence both capabilities and functionings.
Following the above approach, key summary points from each section (dimension) are:
Economic inequality: We analyse this in various ways, including by looking at income and wealth inequality, inequality in living standards, absolute levels of poverty, and employment statistics. This interrogation of the data follows others’ findings that the aggregate Gini coefficient has not changed significantly over time, which we also find to be the case. We further show that within-group racial inequality has increased, especially for Black African and Coloured individuals. In line with others’ findings, we find that most of the inequality is linked to the labour market: both extremely high unemployment and the distribution of wages drive most inequality in South Africa. There is no clear evidence that inequality has improved over time by taking a generational lens. The evidence on whether the inclusion of older generations (>65 years) in income inequality calculations lead to higher inequality is inconclusive. While population growth has been decreasing over the post-apartheid period, it is still higher than in comparable upper-middle-income countries. Population growth also manifests in household structure. Most South African households are extended family households, and these are also the poorest households on a per capita income basis. Single parent households follow with a low per capita income level, as such fertility rates (even though relatively low) may still not be at the desired level for South African women.
Inequality is lower among households with at least one person employed, but only marginally so (relative to the aggregate nationalpicture). Removing the top 1% of households (by income) also does not make a major difference to inequality. Absolute levels of poverty remain high. In 2015, 13.8 million people fell below the food poverty line; and in 2019, 11% of households reported that at least one person went hungry in the last 12 months.
Education inequality: We show that access to education and attendance has improved over time, at all levels. It has in some way reduced inequality in education, but there is still room for improvement. Despite this, the proportion of individuals that complete matric is low (less than half of those who start Grade 1 in a certain year matriculate successfully 12 years later), and this (lower levels of education) has serious implications for chances of employment, as the data show. Most concerning, however, is that the quality of (public) education remains low. It is evidenced by data on outcomes and infrastructure – which often points to the weak conversion of resources into learning outcomes as the main driver.
Health inequality: Like the picture on education, we see that access to health has increased over time, but that this has not translated into meaningful outcomes for beneficiaries due to poor quality of services. As with education, health also does not occur in a vacuum; many other socio-economic factors matter (outside of the health and education system) in working toward outcomes. Overall, health and education inequalities remain high, and follow socio-economic and racial lines.
Living conditions inequality: Like other resources, a clear socio-economic divide exists for housing access. Despite improvement in access to housing over time, including by means of subsidies from the government, many do not have adequate housing. For the 6.8 million households who reportedly have a monthly income of less than R3,500, 64% reported having inadequate housing – defined as informal or traditional houses, dwellings with no flush toilets, and households with more than two people per room. This picture improves slightly for those higher up in the income bracket, but improvement is mainly seen when the household income breaches R15,000 per month (generally the top fifth of households). Access to water, power and sanitation has improved over time, although there is still room for improvement.
Social and cultural inequality: While it is difficult to measure this accurately, we were able to elicit some notable findings. They show to which degree South Africans don’t interact with others who are different in some way, how many South Africans report feeling alienated, how factors such as gender link with poverty, and how South Africans perceive social cohesion to have changed over time.
Physical security and legal inequality: It is clear from evidence on these points that access to physical and legal security, and the impacts thereof, vary significantly by gender, age and geographical location. We discuss the concerningly high statistics of issues such as violence in this section.
Political inequality: Despite South Africa’s democratic system, the voices of individuals are not always heard. According to the data, many also perceive that they cannot discuss their political views freely, and at present, more than 8% of people feel that they cannot vote without feeling pressured.
Environmental inequality: In this final section, we discuss pollution and transport inequalities. Poorer people tend to be significantly more exposed to pollution, often because of residing closer to mines, refuse sites, sewage sites, and main transport roads. We also show the data on inequalities in access to different forms of transport. This indicates the high proportion of people whose mode of transport to health facilities is walking (possibly suggesting transport as a barrier in other areas of their lives, such as in accessing work opportunities).
Overall, we find that the picture of inequality in South Africa is deeply pervasive and entrenched, and it is difficult for many to truly access and participate in society and the economy, which would allow them a chance to improve their own and others’ well-being in their lifetimes.
Despite this, there are examples of progress that deserve to be highlighted. In the past 27 years, South Africa has, for example, expanded its welfare grant programme, housing programme, water and sanitation infrastructure, and access to schooling and healthcare, especially for the poorest and most vulnerable.
Unfortunately, the gains made are insufficient to improve the overall well-being of the majority of South Africans in a meaningful way. Grants are not enough to support recipients, housing and infrastructure programmes are not keeping pace with population growth and urbanisation, and the quality of basic services – including those relating to water, sanitation, health and education – are often suboptimal, resulting in inferior outcomes and services.
Content
Executive Summary
List of figures
List of tables
List of information boxes
Acronyms and abbreviations
1. Background and purpose
2. Conceptual framework – assessing multiple dimensions of inequality
3. Note on data sources
4. Inequality for whom?
5. Economic inequality
5.1 Overall income: inequality and poverty
5.2 Inequality in sources of income
5.3 Post-tax income inequality
5.4 Inequality in post-government expenditure (post-transfer) income
5.5 Inequality in Living Standards
5.6 Nutrition and hunger
5.7 Wealth
6. Education inequalities
6.1 Early Childhood Development
6.2 Basic Education
6.3 Tertiary Education
6.4 How does education link to employment and income?
7. Health inequalities
8. Inequalities in living conditions
8.1 Housing
8.2 Electricity, water and sanitation
8.3 Digital connectivity
9. Social and cultural inequalities
10. Physical security and legal inequalities
11. Political inequalities
12. Environmental inequalities
12.1 Pollution inequalities
12.2 Transport inequalities
13. Linking the many dimensions of inequality: Conclusion
References
List of figures
Figure 1: Framework for assessing multiple dimensions of inequality using a capability approach[2]
Figure 2: Growth in the number of individuals and households in South Africa, 2001-2019
Figure 3: Household sizes and differences in the measurement of income across household surveys, NIDS 2014/2015 vs. Living Conditions Survey (LCS 2014/2015)
Figure 4: Average annual percentage population growth, 1985 – 2020
Figure 5: Distribution of households by household structure type (total SA population)
Figure 6: Personal income tax contributions in 2018
Figure 7: Racial composition of post-tax income groups, 2019
Figure 8: Total individualised transfers received by pre-tax income group
Figure 9: Households' main income source by LSM group
Figure 10: Reported levels of household hunger (17.2m households)
Figure 11: Household wealth (Rand billion) according to data of the South African Reserve Bank (2020)
Figure 12: More accurate estimate of household wealth (2020, Rand Billion)
Figure 13: Residential properties on the deed's office registry by value
Figure 14: Linkages to the labour market and access to financial assets (wealth) through employment for the economically active
Figure 15: Educational attendance by age and reasons for non-attendance
Figure 16: Percentage distribution of education attainment for individuals 20 years and older, 2002-2018
Figure 17: Share of population by highest level of education and age group, 2020
Figure 18: Share of population aged 25-64 by highest level of education and race, 2010 and 2020
Figure 19: Demographic shift in young graduate (<35 years) profiles by race and gender between 2008 and 2020
Figure 20: Proportion of the unemployed by education level
Figure 21: Narrow rate of unemployment by highest level of education
Figure 22: Educational profile (highest level of education of the economically active population, by race
Figure 23: Relationship between highest level of education and unemployment rate for the economically active, by race (2021)
Figure 24: Narrow rate of unemployment for young graduates (2008-2020)
Figure 25: Means of transport to health facility by geographic location (total households=17.2m)
Figure 26: Reason for not using closest healthcare facility (by geographic location)
Figure 27: Medical scheme membership by race group
Figure 28: Number of households living in different types of housing, 2001-2019
Figure 29: Number of households living in different types of housing, 2015/2015
List of tables
Table 1: Mid-year population estimates 2021 by province
Table 2: South Africa's age distribution
Table 3: The Gini coefficient for the period 2007-2019 calculated from various data sources (tax and household surveys)
Table 4: Comparing the Gini coefficient for all households vs. households excluding the top 1% of income households
Table 5: Gini coefficient for individuals aged 65 and younger vs. individuals older than 65
Table 6: Inter-group Gini coefficient for those aged 65 and younger (excluding those aged older than 65)
Table 7: Median and average monthly per capita incomes of households, by household structure
Table 8: Gini coefficient for employed households, all individuals and all households
Table 9: Gini coefficients by population group for households with at least one member employed
Table 10: Number of registered taxpayers vs actual number of tax payments received, 2016-2020
Table 11: Afrobarometer responses about a survey on voting freedoms without pressure
Table 12: Different fuel sources used for cooking, by population group, 2003 & 2019
Table 13: Population groups' modes of transport to healthcare facilities
List of information boxes
Info Box 1: Impact of the COVID-19 pandemic on social grants
Info Box 2: Living Standards Measure
Info Box 3: Patient treatment journey
Info Box 4: Gugulethu street committee leader
Info Box 5: Pollution near coal mines
Acronyms and abbreviations
ECD Early Childhood Development
GBV Gender-Based Violence
GEMS Government Employee Medical Scheme
GHS General Household Survey
IES Income and Expenditure Survey
LASA Legal Aid South Africa
LCS Labour Conditions Survey
LSM Living Standard Measure
MDR-TB Multidrug Resistant Tuberculosis
NIDS National Income Dynamics Study
NIDS-CRAM National Income Dynamics Study – Coronavirus Rapid
Mobile Survey
OHS October Household Survey
QLFS Quarterly Labour Force Survey
SAARF South African Audience Research Foundation
SARS South African Revenue Service
StatsSA Statistics South Africa
TERS Temporary Relief Scheme
TIMSS Trends in International Mathematics and Science Study
1. Background and purpose
“‘Inequality’ has been likened to an elephant: you can’t define it,
but you know it when you see it.” (Fields, 2002)[1]
Almost three decades have passed since the fall of apartheid, and income inequality persists. Economic growth, equity and social cohesion in South Africa are generally considered poor or regressive despite considerable policy reform aimed at improving these outcomes. The media and political publications continue to place emphasis on income disparities using aggregate income inequality measures, such as the Gini coefficient. While important, this seemingly one-dimensional view of tracking progress post-1994 risks missing out on the dimensions of inequality that could become levers of change if properly understood.
Strides have been made in terms of achieving equity in access to services – particularly in the areas of education and health – and these should be acknowledged. At the same time, these strides have not translated into meaningful shifts in outcomes. It is therefore critical to understand the drivers of inequality to ensure that government policies and civil responses achieve the most optimal impact.
A shift towards a multidimensional framing of poverty and inequality in South Africa lends itself to a more comprehensive description of what has – and has not – been achieved during the new democratic dispensation. By interrogating the changes made in South Africa in a more granular way, the lived experience of South Africans may better be understood and reflected. The approach also allows the drivers of inequality to be identified. In this way, levers of change and opportunity can be described, and poverty can be tackled more comprehensively.
This report takes a deeper look at multiple dimensions of inequality and how it has changed in South Africa over the last 27 years. Priority pathways out of poverty and inequality traps are described in a separate report.
2. Conceptual framework - assessing multiple dimensions of inequality
We use various dimensions and metrics to describe and measure inequality. For the purpose of this report, we will be evaluating inequality based on the capability approach taken to the multidimensional inequality framework (explained later in the text) typically used in the broader Human Development context.[2] While the exact dimensions of any multidimensional framework can vary somewhat, we opted to interpret typical dimensions presented in the literature in a way that allows for more concrete analysis given limitations on data availability.
“Through the capability approach the Human Development approach redefines
the concept of well-being instead of on survival means.” (Bucelli and McKnight, 2021)[2]
Figure 1: Framework for assessing multiple dimensions of inequality using a capability approach[2]
The framework provides a systematic approach to evaluating eight dimensions of inequality, while emphasising the interconnection between all dimensions.[2] These interconnections are based on Sen’s capability approach. The capability approach distinguishes between conversion factors (drivers of multidimensional inequality), capabilities and functionings.[2] Conversion factors influence the degree to which advantage or disadvantage can move between the individual domains of inequality, some of these being capabilities (e.g. learning and education) and other functionings (outcomes – e.g. health). An environmental factor such as pollution is an example of a conversion factor that can drive the degree to which inequalities are transferred between domains – for example, from health to economic, if individuals are no longer able to work due to ill health because of pollution exposure.
The progress (or lack thereof) made in each dimension of the multidimensional inequality framework will be explored using a variety of metrics to provide a comprehensive understanding of inequality in South Africa.
3. Note on data sources
Our analysis makes use of various data sources (household surveys, public sector administrative data sources and private sector data sources) to ensure that our evaluation is as comprehensive as possible. Furthermore, this approach ensures robustness since each data source has its own strengths and weaknesses.
The underlying data used in this report and quoted in other reports is mainly derived from survey data sources. In some cases, it is possible to calculate income (personal, household and/or per capita income) from these surveys. However, incomes may be poorly reported in surveys – in some cases because respondents deliberately misstate their income and in others because incomes are inherently unstable and difficult to report on. For this reason, many researchers prefer to use expenditure data rather than income data for analysis. At the same time, the household unit itself is not well-defined or stable, which further complicates deriving per capita income measures.
4. Inequality for whom?
The starting point for any discussion on inequality is defining the population for whom inequality is being measured. An inequality assessment presupposes a distribution of resources relative to a given population within which the ranking of resources relative to individuals can be done. According to the 2019 General Household Survey (GHS), the most recent version of the survey for which data is available, there are 58.4 million people and 17.2 million households in South Africa. The number of households in South Africa grew at 2.4% per annum while the number of individuals grew at 1.5% per annum between 2001 and 2019. This discrepancy is driven to a large degree by the continued growth in one-person households, a phenomenon commonly associated with urbanisation, and in South Africa, with migration and employment patterns. The growth in households has clear implications for service delivery; an increase in the number of households implies a need for more dwellings, albeit smaller, and more infrastructure to service these dwellings appropriately.
Figure 2: Growth in the number of individuals and households in South Africa, 2001-2019
Sources: Census 2001, Community Survey 2007, Census 2011, Community Survey 2016, GHS 2019
Growth in population and households is also not evenly distributed across the country, with noticeably higher population growth in better-developed areas, as individuals migrate to access jobs, better-quality housing, education and healthcare. The latest mid-year population estimates published by Stats SA in July 2021 indicate that Gauteng has seen net in-migration of almost one million people between 2016 and 2021, with the Eastern Cape, Limpopo and Kwa-Zulu-Natal experiencing net out-migration of 320,000, 189,000 and 84,000 people respectively.
Table 1: 2021 mid-year population estimates by province
Source: Stats SA 2021
Recent statistics also show that 10% of the national population was aged 4 and under, 25% aged 12 and under, and 50% aged 26 and under (see Table 2 below).
Table 2: South Africa's age distribution
Source: NIDS Wave 5 (2017)
5. Economic inequality
When viewing inequality through a capability lens, income and wealth are not ends in themselves but serve as means to achieve welfare and freedom.[3] Because this report uses a multidimensional inequality framework underpinned by a capability approach, income and wealth inequality in South Africa is explored from the perspective that it contributes to suboptimal functionings among those at the bottom end of the distribution.
Income can be considered anything that makes consumption possible[4] and therefore, its sources are vast and varied. However, it is useful to categorise income according to its factor source (the factor of production used to generate the income): labour income (wages and salaries), capital income (capital gains and interest), income from entrepreneurship (profit) and land (rent). Understanding the dynamics and structural nature of the factor markets that generate these incomes is essential to figuring out the inequality story. Appreciating these complexities of factor markets also helps with accepting that income inequality is ‘sticky’ (slow changing).
Another important source of income in the South African context is transfers from the government in the form of social security grants. Spending on income grants and taxation are fiscal tools used by governments to improve distributive outcomes in the short term. One way to determine the impact of these direct interventions in reducing income inequality is by measuring primary (pre-tax and pre-government spending) income and comparing it to post-tax secondary income (post-tax and post-government spending). Although it’s more difficult to measure than income, we also consider the distribution of wealth in South Africa because of its potential to generate longer-term intergenerational benefits.
5.1 Overall income: inequality and poverty
South Africa has been labelled as one of the most unequal countries globally for decades. This is due to the strong focus placed on the Gini coefficient as the primary measure of inequality, which has remained persistently above 0.6.[5] The 2019 Stats SA inequality report[6] that reviewed progress in inequality reduction in post-apartheid South Africa arrived at the same conclusion. It revealed that there had been no substantial change in South Africa’s Gini coefficient during the post-apartheid period.
Even though the report[6] also considered non-income dimensions of inequality and numerous measures of inequality, South African media and political organisations focused exclusively on the aggregate Gini as a summary measure of inequality. Throughout this report, we will draw attention to improvements in inequality in other dimensions, where it exists. But ultimately income inequality – as measured by the Gini, in particular – remains the most widely reported and considered measure when assessing post-apartheid inequality.
Differences in both the reporting and consequently ultimate measurement of income reported by different households can make it difficult to accurately track income inequality. In the 2014/2015 Living Conditions Survey, one of the biggest household surveys in South Africa, 10% of households reported earning income of R30,000 or more per month. In the NIDS survey, a smaller panel survey conducted in the same period, only 6% of households reported earning income of R30,000 or more. Likewise, at the other end of the market, according to NIDS 2014/15, 38% of households earn less than R3,500 per month, compared to 41% as reported by the Living Conditions Survey. The Gini coefficient for these surveys is 0.6.
Figure 3: Household sizes and differences in the measurement of income across household surveys, NIDS 2014/2015 vs. Living Conditions Survey (LCS 2014/2015)
Sources: NIDS 2014-2015, Wave 4; LCS 2014-2015
Nevertheless, the reporting errors in and between surveys are typically not large enough to alter the pattern of extreme inequality shown by survey data. For example, while the NIDS, LCS, GHS and SARS tax statistics all provide markedly different estimates for the Gini coefficient, they leave no doubt that South Africa is characterised by extreme income inequality (Table 3). The Gini coefficient is lowest when administrative tax data is used (varying between 0.41 and 0.50 between 2007 and 2019). Data from the GHS shows Gini coefficients between 0.56 and 0.62, while the NIDS data (using a smaller sample) has Gini coefficients varying between 0.61 and 0.7. The LCS is not regularly repeated.
Table 3: The Gini coefficient for the period 2007-2019 calculated from various data sources (tax and household surveys)[1]
Some critical data sources, such as the Quarterly Labour Force Survey (QLFS), do not include sufficiently useful income measures (household income). Even where personal income data is collected from the respondent, this data is not disseminated publicly.
Where this is the case, the analysis uses other proxies to segment the market. Examples of these alternative proxies include Living Standard Measures (or LSMs, which themselves reflect key dimensions of inequality such as access to services), area or area type, race as well as education levels. The unit of analysis relative to which inequality is calculated varies throughout the report.
Excluding the top 1% from income inequality calculations only marginally reduces overall inequality. The contribution of the top 1% of the income distribution is a topic of global relevance and concern.7 Removing the top 1% of the income distribution’s income from calculations of the Gini coefficient can indicate to what extent overall income inequality is driven by the phenomenon of the “top 1%”, as the literature often refers to this group. As shown in Table 4, excluding this group’s income from the Gini only marginally decreases overall inequality. It generally lowers the Gini from 0.60 and above to just below, although using GHS income data lowers it to as low as 0.52 – which is still a very high level of inequality. Given that good data on the total population’s wealth is not accessible, this exercise was not repeated with wealth data.
Table 4: Comparing the Gini coefficient for all households vs. households excluding the top 1% of income households
Sources: NIDS and GHS
The evidence on whether the inclusion of older generations (>65 years) in income inequality calculations lead to higher inequality is inconclusive. Older generations have theoretically had more time to accumulate wealth from non-labour market income in the form of both financial and non-financial assets such as housing. However, different older population groups in South Africa would also have been more heavily subject to the inequalities generated by apartheid. Depending on the data source, the Gini coefficient for the population aged 65 and below can be higher or lower than the Gini coefficient for the population older than 65 years (Table 5). Data from the GHS generally shows a marginally lower trend in inequality for the population aged 65 and younger, while data from NIDS shows higher inequality among the population 65 and younger. The inter-group Gini coefficient for different South African race groups (Table 6) shows much higher inequality among Black and Coloured South Africans compared to White and Indian/Asian South Africans. Specifically for Black South Africans aged 65 years and younger, there also was a trend of increasing inequality in later years. Among Coloured South Africans, the inter-group Gini coefficient seems to have decreased between 2015 and 2019. This raises questions about which income distribution brackets within the group experienced increased or decreased incomes. Further investigation is required to answer this question.
Table 5: Gini coefficient for individuals aged 65 and younger vs. individuals older than 65
Sources: NIDS and GHS
Table 6: Inter-group Gini coefficient for those aged 65 and younger (excluding those aged older than 65)
Source: GHS
South Africa has experienced a population growth of between 1.2% and 2.3% over the past 36 years. The population growth rate has declined from an average of 2.2% per annum between 1985 and 1990 to an average of 1.2% per annum between 2015 and 2020. The country’s average population growth rate has typically been between the average of upper-middle-income countries’ growth rate and the lower-middle-income countries’ growth rate (Figure 4). Kenya, a comparator African country (but falling into the lower-middle-income category), has experienced a higher average population growth from 1985 to 2020, decreasing from 3.5% to 2.5%. All middle-income countries have experienced declining population growth rates over the period. This is likely due to the increased economic development of these countries and increasing access to education, family planning and other similar services.
Figure 4: Average annual percentage population growth, 1985-2020
Source: United Nations, Department of Economic and Social Affairs, Population Division, 2018
Population growth may play out in household structures (in an inequality context). Although fertility levels in South Africa (which contribute to decreasing average population growth rates) have been declining significantly, suggesting a strong fertility transition[8], average fertility may still be higher than women’s and households’ stated preferences.
Stats SA recently (2020) published nationwide research in which they asked women who had a baby in the five years preceding the survey (conducted in 2016) whether they wanted to get pregnant at the time. If they answered no, they were asked whether they wanted to rather have a baby later (indicating that the birth may have simply been unplanned or mistimed) or did not want more children (therefore indicating that the birth was unwanted). Unwanted births are therefore births – as recalled by the women surveyed – where no additional birth was planned or wanted at the time of conception. Over 20% of total births from the prior five years were classified as unwanted according to this research. The underlying research also shows that another 34% of the births were mistimed. As we discuss below, this may also have implications, especially for younger mothers. These unwanted births show a decline with increasing education; in 2016 unwanted births to mothers with tertiary education (11%) was four times less compared to mothers with no education (46.3%). 26% of the unwanted births were birthed into households in the poorest wealth quintile and 13% into households in the richest wealth quintile. The highest number of unwanted births were found in the Eastern Cape, followed by KwaZulu-Natal and Mpumalanga.[9]
Various factors may contribute to these high levels of unwanted fertility, including low-quality contraceptive services in the public sector, and gender violence and inequality. Recent research indicates that the unmet need for contraception remains high – at 19% – for sexually active women, and at 15% for married/in-union women. This research also highlights the widespread dissatisfaction by both community members and healthcare providers with the level of family planning quality of care.[10] We discuss gender-based violence later in this report under physical security and legal inequalities, where it is found that among adult women in South Africa, 21% have reported experiencing physical abuse in their lifetime.
While ideal population growth rates may be ambiguous and even controversial, the findings from stated preferences do provide insights into the possible links between population growth, poverty and inequality. The same report regarding unwanted births by Stats SA discusses that the disadvantages suffered by unwanted children may be in their health, early childhood development and potential future social and economic opportunities.[9] Similarly for parents (especially mothers), unwanted births may, dependent on the mother’s age at birth, limit one’s education, and job and income prospects, with possible knock-on effects for many other aspects that perpetuate inequalities. It is notable that there is also two-way directionality here (fertility influencing income and income influencing fertility), with other studies finding that higher education and income lead to lower fertility levels.[11]
Teenage births may have declined over time, but StatsSA shows that in 2020 alone, approximately 34,000 births were to women aged 17 and younger. These births are likely to cause education interruptions and later entry into the labour market. In the same year, it was found that more than 60% of all births were registered without the details of the father. This percentage may change alongside legislative changes and there may be many reasons for this phenomenon, including choices by individuals not to marry (until now prohibiting the father’s details to be included on the birth certificate).[12] Nevertheless, it does raise the question of the prevalence of single parents and what implications this may have for poverty and inequality.
In relation to inequality, population growth may indeed play out in household formation and different types of households’ average income and poverty levels.
Figure 5 sets out the distribution of different types of households in South Africa in 2015, with 39% of South Africans living in extended family and/or non-related households and 11.9% of households being single-parent family households. Extended family/non-related households typically have larger sizes than the other households. Table 7 sets out the income per capita of each household structure. The median per capita monthly income of the extended and/or non-related households is the lowest of all household structure types at R1,283 per month, followed by single-parent family households at R1,462 per capita per month.
Population growth which manifests in extended family/non-related households and in single-parent families (often female-headed) contributes to inequality in that these households typically share a limited set of resources among a large group of people, tend to be grant reliant, and these types of households typically don’t have good labour market links.[13] This type of household formation is both a result of poverty and inequality, as well as a contributor to future inequality.
Figure 5: Distribution of households by household structure type (total SA population)
Source: Living Conditions Survey 2015
Table 7: Median and average monthly per capita incomes of households, by household structure
Data source: Living Conditions Survey 2015
COVID-19 has worsened the economic situation and will affect poverty and inequality. Given the pre-pandemic South African context of weak, non-inclusive economic growth and high unemployment rates, it is expected that the economic effects of COVID-19 may perpetuate existing income inequalities and poverty. In some cases, it may even reverse economic gains made amongst vulnerable groups over the past two decades (e.g., women, the previously disadvantaged). Recent studies already point in this direction. Findings from the nationally representative NIDS-CRAMc survey provide robust evidence of sharp increases in household and child hunger and insufficient money for food during the COVID-19 pandemic, both of which remained disturbingly high from May 2020 to May 2021.14 When considering the impact of the COVID-19 crisis on the labour market - the source of labour income - gendered effects are already emerging. By March 2021, men’s employment and working hours reverted to pre-COVID levels but in contrast, women’s were below February 2020 baseline figures. Furthermore, inequalities in the time spent on childcare and in the income support for unemployed or furloughed workers endured during COVID-19.[15]
When Statistics South Africa upper-bound poverty line is used, more than one out every two South Africans were poor in 2015.[16] While there had been a clear decrease in poverty (using the upper-bound poverty line) between 2006 and 2011 from 66.6% to 53.2% of the population, by 2015 poverty had increased to 55.5% of the population.[16] This meant that 30.4 million South African were living in poverty in 2015. A similar poverty trends report has not been produced by Statistics South African since 2017 but the impact of COVID-19 is likely to have exacerbated the poverty situation by a large quantum.
5.2 Inequality in sources of income
Since the end of apartheid, a divergence between the top and bottom income deciles in real factor incomes (income generated through labour, capital, land or entrepreneurship) was seen.[17] The 13% increase in national income was largely due to a 30% increase among the highest decile, with an almost 50% increase seen in the top 1%.[17,18] In contrast, the average factor income has remained static in the middle 40% group, while the lower 50% has had a 30% drop.[17,18]
Economic growth since the 2000s has primarily resulted in an increase in income among the high-income earners.[17,18] Increases among the lower 90% have failed to rise substantially, and returned to levels observed in 1993 during 2011 (on the back of the global financial crisis). However, the top 10% was largely unaffected during 2011.[17,18]
Labour market income and large variation in labour market income play a very significant role in overall inequality. Work done by Stats SA and the French Development Agency found that roughly two-thirds of total inequality derives from inequality in labour market earnings and that half of this is related to the very high levels of unemployment in South Africa. The distribution of wage earnings among the employed also shows a long upper tail which increases overall inequality.
In this report, a simple exercise was done to determine the relationship between inequality and access to labour market earnings. Inequality is lower amongst households with at least one employed member, but only somewhat (see Table 8). Compared to data presented for the total population earlier in this report, inequality among households with at least one employed member is only marginally lower than for the total population. This could be because having only one employed member is an insufficient threshold, also given the high variation and longer upper tail described in the income distribution. Using household survey data that contains only household income data (as opposed to personal income) limits the analysis that can be done to households where at least one member is employed. If income is spread between many household members, this reduces per capita income, but total income would still be higher than in households without any employment.
The same pattern of lowest inequality within the groups of White and Indian/Asian South Africans hold here, with the highest inequality among Black and Coloured South Africans.
Table 8: Gini coefficient for employed households, all individuals and all households
Sources: GHS and NIDS
Table 9: Gini coefficients by population group for households with at least one member employed
Source: GHS
5.3 Post-tax income inequality
Personal income tax contributions, a large source of tax income that enables government expenditure, are paid primarily by wealthy South Africans because of the progressive nature of the tax regime. As illustrated in Figure 6 below, the top quartile of earners contributed 73% of the total personal income tax collected in 2018. In 2018, there were about 22.2m registered taxpayers. 19.1m tax payments were received (Table 10), although only 5.9m taxpayers were assessed.
Figure 6: Personal income tax contributions in 2018
Source: SARS data, 2018
Table 10: Number of registered taxpayers vs. actual number of tax payments received, 2016-2020
Source: SARS data, 2016-2020
Despite a progressive personal income taxation regimed and substantial changes in service delivery in the post-apartheid period as reviewed elsewhere in this report, racial inequality persists in South Africa. The White-to-Black income ratios remain high at 8 overall, and this divide is even greater when considering wealth.[17] A basic picture of the receipt of post-tax income by population share in 2019 is presented in Figure 7. In this figure, the population has been ranked by income from the lowest to the highest income receiving. After accounting for transfers and taxes, White South Africans account for a large proportion (50%) of the top 1%, indicating that they still receive far more income than their population share.[17,18] The slight reduction in racial inequality over the last two decadeshas largely been driven by the top 1% who are Black African.[18] The increase in incomes of Black South Africans falling in the top 1%and other higher-income groups have, however, increased within group inequality among Black South Africans.
Figure 7: Racial composition of post-tax income groups, 2019[18]
5.4 Inequality in post-government expenditure (post-transfer) income
Post-transfer income refers to income after government transfers (like social security) and expenditure on social services (education and health) have been taken into consideration. Measuring post-transfer incomes allows one to assess the impact of government transfers on primary income (pre-tax and pre government spending income), and whether it is progressive or regressive.
Figure 8 below shows the share of total transfers in grants, education and health that have been allocated to various income groups from 1993 to 2019.[18] It depicts the progressive changes made by the South African government: Individualised transfers as a proportion of the national income have consistently increased in favour of the poor.
There has been a relatively rapid growth in transfer income among the bottom 50% who received approximately 12% of the national income.[17,18] The middle income group also experienced an increase in transfers from 3.9% in 1993 to 5.3% in 2019.[17] This is compared to the 1% of national income received by the top 10%.[17] Overall, transfers have been made in the form of cash transfers or in-kind transfers.[17]
Figure 8: Total individualised transfers received by pre-tax income group[18]
Info Box 1: Impact of the COVID-19 pandemic on social grants The COVID-19 pandemic has been a driver of the rapid expansion of social grants in South Africa. This was due to the increase in unemployment. The recovery response indicated that R50 billion would be used to provide support to those who are vulnerable, along with the distribution of vouchers and food parcels.[15] The COVID-19 recovery response has resulted in the following grant adjustments or additions:[16]
R16.5 billion in relief was also provided to businesses in the form of debt holidays and financial support.[15] |
5.5 Inequality in Living Standards
It is possible to consider economic inequality in terms of more than just income. One can also view it in terms of what people choose to buy with their earnings – as well as their everyday living standards, as captured by a measure that not only takes income into account, but also assets and access to services. The South Africa Audience Research Foundation’s (SAARF) Living Standards Measure aims to do precisely this (see Info Box 2). In a way, it’s a multidimensional economic measure.
Info Box 2: Living Standards Measure In the 1980s, SAARF sought to construct a tool that could be used for market research.[17] Multiple variables were combined to create an index called the Living Standards Measure (LSM). This consisted of variables that were found to be strong indicators of living standards in South Africa based on the results of the AMPS survey.[17] These indicators could be used to segment the population into 10 LSM groups.[17] After multiple iterations of the LSM, 20 indicators were incorporated in 2001.[17] These evaluated whether households have the following: a built-in kitchen sink, car, flush toilets, shopping at supermarkets, microwaves, credit cards, fridge, washing machine, financial services, hut, stove, polisher, insurance policy, Hi-Fi, video cassette recorder, domestic worker, TV, car, hot running water and telephone.[17] The list of relevant assets has been updated with time as certain assets have lost their economic and functional significance. Each variable has a positive or negative weighting used to determine the final index.[17] |
The 2014-2019 period saw households from the LSMs 6 and 7 become more reliant on grants and remittances as their main sources of income. Households in the highest LSMs rely largely on income through salaries, business profits and pensions, and the latter two have become more prominent in the past five years. More than 25% of households at LSM 5 and lower rely on grants as their main source of income. Remittances also play an important role for households in the lower LSMs (approximately 15% of these households’ main source of income), particularly in more recent years.
Figure 9: Households' main income source by LSM group
Source: SAARF
5.6 Nutrition and hunger
Hunger is a sign of insufficient income to sustain good nutritional levels. It is closely linked to income, wealth and living standards. However, we treat it separately here, given its importance as an indicator of insufficient means to sustain a basic human need. In the latest General Household Survey (2019), 11% of all South African households reported at least one person going hungry in the past 12 months. 6% of households reported sometimes going hungry, while 2% said they often go hungry. These percentages are much higher for Black South African households, where 12% of households reported at least one person going hungry in the past twelve months, compared to less than 1% of White South African households. Hunger levels in South African households were exacerbated by the COVID-19 pandemic and associated lockdown measures, which prevented many households from obtaining an income and ultimately, necessary levels of nutrition.[14]
Figure 10: Reported levels of household hunger (17.2m households)
Sources: GHS 2019
