Content uploaded by Joshua Rosenberg
Author content
All content in this area was uploaded by Joshua Rosenberg on Jun 23, 2022
Content may be subject to copyright.
Reconciling ecological, social, and economic development in South
Africa: a strong sustainability perspective
Joshua ROSENBERG
Supervisors: A. GODIN† & A. DAVID‡
† godina@afd.fr, Agence française de développement, Centre de recherche en économie de l’Université
Sorbonne Paris Nord, France
‡ davida@afd.fr, Agence française de développement, South Africa
Economic Policies for the Global transition (EPOG+) – Joint Master Degree
Major C — Development, sustainable development, and the ecological transition
June 2022
2
This thesis was written in partial fulfilment of the requirements of the one-year track of the
Economic Policies for the Global Transition (EPOG) joint master degree conferred by
Sorbonne Université, Université Paris Cité, and the Université de Technologie de
Compiègne. Supervision was graciously provided by Antoine Godin and Anda David,
associated with the Agence française de développement. The opinions, results, and validity
thereof are those of the author alone, with whom responsibility for all errors lie.
Abstract
This paper attempts to uncover the possible futures of development of the environment,
social, and economic spheres in South Africa. A history of GHG and CO2 emissions, social
protection coverage and expenditure, and the labour market are provided to contextualise
these areas, as well as their recent developments and trends. Further, a methodological
framework centring a strong sustainability approach is employed, assessing what severity
of constraints, given two scenarios, exist in achieving goals of emission reduction as per
the September 2021 NDC targets, with regard to fulfilling constitutional obligations in
providing adequate social protection coverage, unemployment targets as per the NDP,
public debt levels, and the balance of payments. Results show that with current energy
generation and productive structures, harsh constraints make meeting these goals unlikely.
However, we also show the mechanisms by which these constraints may be relaxed.
Ultimately, it will fall to political will, succinct industrial policy, and investment in order
to enable any meaningful ecological transition.
Key words: Sustainable development; Strong sustainability; Development trajectories
3
Economic Policies for the Global transition (EPOG+)
Plagiarism declaration
I, the undersigned,
Last name: Rosenberg
First name: Joshua
Cohort: 2021/22
Major/minor: C3 - Development, sustainable development, and the ecological transition
One or two-year programme: One
Title of Master’s thesis: Reconciling ecological, social, and economic development in South
Africa: a strong sustainability perspective
Name of supervisor 1 (Paris): Antoine Godin
Name of additional supervisor: Anda David
Hereby formally declare that I have written the submitted Master's thesis entirely by myself
without anyone else's assistance. Where I have drawn on literature or other sources, either in
direct quotes, or in paraphrasing such material, I have referenced the original author or authors
and the source in which it appeared.
I am aware that the use of quotations, or of close paraphrasing, from books; magazines,
newspapers, the internet or other sources, which are not marked as such, will be considered as an
attempt at deception, and that the thesis will be graded as a fail. ln the event that I have I have
submitted the dissertation – either in whole or in part- for examination within the framework of
another examination, I have informed the examiners and the board of examiners of this fact.
Place: Paris, France
Date: 30 May 2022
Signature:
4
Acknowledgements
I have received support from many people over the course of writing this thesis. I wish to
acknowledge those people and extend my gratitude towards them.
Firstly, to my lovely supervisors, Antoine Godin and Anda David. Thank you both for your
support, patience, and trust. Thank you also for the dedication of your time and efforts in very
early morning meetings. I hope these efforts were worth it!
To Ilan Strauss, for taking me along for a great ride during COVID-19. I was able to see what
futures in economics were possible, without which I would likely not have pursued my studies
further. For this, I owe you a great debt (rates negotiable). I hope this serves as one small step in
closing the Grand Canyon of talent and experience between us.
To David Flacher, Catherine Rolland Landy, and the whole EPOG consortium team. Thank you
for your tireless efforts and support in making this programme possible in the way that it is.
Heterodox-aligned courses and programmes are few and far between, though desperately wanting.
To my friends, home and abroad, for holding my hand, making me laugh, and setting me straight
when needed. To Fatmanur, Kumbz, Kemeel, Gianna, Selma, Heloisa, Aliya, and Kamal, thank
you for the special moments shared. To my EPOG colleagues, thank you for many great
discussions.
To Eleni and family, for sharing your lives with me.
To my family, those here and passed, thank you for keeping me in good spirits, laughing
hysterically at my memes, and for your loving support and confidence in me. To the Eds, bless.
Finally, to Daniel Dumile and Jun Seba, for the constant company and inspiration. Rest in peace.
5
Table of Contents
1. Introduction ............................................................................................................................. 6
2. Review of literature ................................................................................................................. 8
3. Historical Trends in Emissions, Social Protection, and Employment ...................................... 11
3.1 Environment and emissions ................................................................................................ 11
3.2. Social protection coverage ................................................................................................. 21
3.3. Employment and labour ..................................................................................................... 30
4. Methodology ............................................................................................................................. 39
5. Results and Discussion ............................................................................................................. 42
5.1. Emission reduction and output perspective ........................................................................ 42
5.2. Employment perspective .................................................................................................... 45
5.3. Social protection spending and macroeconomic linkages.................................................. 49
5.4. Public debt constraints and dynamics ................................................................................ 52
5.5. Balance of payments constraints and dynamics ............................................................. 55
5.6. Political implications and considerations ....................................................................... 57
6. Limitations and further work ................................................................................................. 58
7. Conclusion ............................................................................................................................. 60
Reference List ............................................................................................................................... 62
Appendix ....................................................................................................................................... 67
6
1. Introduction
South Africa faces an enormous challenge in the restructuring of relations after the dawn of its
democracy in 1994. An economy built on resource extraction and labour exploitation, a fragile
social compact, and a long history of racial oppression required large efforts, politically and
economically, to bring South Africa into a sunnier state of affairs.
Whilst there have been noticeable improvements in many facets, South Africa remains in the midst
of crises of poverty, inequality, and unemployment. Whilst these ailments are symptomatic of an
obviously unhealthy history, a newer comorbidity enters the fray in the form of the need to reduce
carbon emissions. With a history of mineral extraction and a decrepit national energy generation,
transmission, and distribution system, the increasingly entangled mineral-energy-complex (MEC)
through financialisation increases the severity of this problem, as well as that of enacting change
in remedying it (Ashman, 2021).
To contextualise the need for a transition away from coal-powered energy generation, it is worth
noting that South Africa produced more than twice the amount of emissions than any other African
country (Statista, 2022), though holds only 4.5% of the continent’s population.1 With some 87%
of production-based emissions being from the use of coal, it is clear that South Africa requires a
rapid and broad restructuring of its energy generation system. South Africa updated its Nationally
Determined Contributions in September 2021, aiming for net CO2 emissions of between 398-510
Mts of CO2 equivalent by 2025 and 350-420 Mts by 2030 (Presidential Climate Commission,
2021). From 2019 levels, this implies decreases of 23.5% and 33% by 2025 and 2030 respectively.
On the consumption side, South Africa’s intense income inequality is reflected in its emissions,
with the top 4% consuming 38.4 tons of CO2 per capita per year, compared to 0.3 tons for people
in the bottom 20% (Arndt et al., 2013). The embattled state-owned energy giant, Eskom, has
recently seen some support in the form of bilateral financing of US$8.5 billion announced at
COP26 towards a future free from coal dependence (UN Climate Change Conference UK 2021,
2021). However, major concerns remain over the appropriateness of this financing measure.
Unemployment rates in South Africa are high by any measure, being 34.9% at last count (narrow
definition), rising to 46.6% in the expanded definition (Statistics South Africa, 2021a). The reasons
for this are many, but include the economic structure of the economy inherited from the Apartheid
era, geographic disparities, declining public and private sector investment, and dwindling (and
negative) sectoral growth. Addressing the crises of poverty, inequality, and unemployment hinges
largely on employment and GDP growth. The means by which this can be improved have not
borne fruit for the South African people since at least the end of the commodity supercycle.
Considering these challenges, many people require assistance from the state. There has been a
rapid increase in the number of those benefiting from various social assistance programs over the
1 Whilst large in the African context, this equates to only 1.3% of total global CO2 emissions.
7
last 2 decades, from 3.9 million people in 2000 (9% of the population) to 18.3 million people (31%
of the population) (Business Tech, 2021). Predominantly, these are people below the age of 18 and
people above the age of 59, corresponding to the Child Support Grant and the Old Age Grant
(South African Social Security Agency, 2021a). However, prior to COVID-19, there was almost
no state support offered for those between the ages of 18 and 59.2
After sustained pressure from civil society, a R350 (~US$23) per month grant was announced for
those with no other income (Ramaphosa, 2020). This, though not without its issues, provided a
lifeline to nearly 10 million people. Further, more attention was brought to this existing gap in
South Africa’s social security net. Academics, various civil society organisations, and indeed even
the Department of Social Development have put forward proposals to close this gap in the form of
a Universal Basic Income Guarantee (UBIG) (Institute for Economic Justice, 2021). Whilst many
modelling scenarios and options have been considered, the introduction of a UBIG; the closing of
the gap in the social security system requires adoption and stamp of approval from the office of
the Presidency as well as National Treasury. This has proven to be difficult, however, due to their
neo-classical understanding of the economy paired with ongoing attempts at deepening fiscal
consolidation (austerity) on the expenditure side (Isaacs, 2020). In the context of damning socio-
economic indicators, this gap in the social security net becomes an increasingly worrying issue,
denying millions the chance to live with dignity. Some progress has been made in this regard,
however, with civil society organisations and academics having recently met with President
Ramaphosa, Minister of Social Development Zulu, and Finance Minister Godongwana around the
issue (Mafata, 2022).
Much has been written on the challenges of poverty, inequality, unemployment, and social security
net gaps. Emission reduction holds less space in South African political and economic discourse,
though increasing. In light of an approach of strong sustainability, these challenges need to be
addressed simultaneously. A lack of attention and funding not only decreases the efficacy in other
areas, but leads to an unstable and unsustainable development path. Indeed, these challenges are
not all inherently reliant on economic growth. Carbon emission reductions specifically should not
be addressed in a growth-centric manner. However, increasing aggregate demand through
increased employment will increase output, from which increased tax revenues are traditionally
used to finance increased social security measures. South Africa, as with many other African and
developing countries, is not afforded the luxury of abandoning economic growth.
With a litany of environmental, social, and economic goals to achieve, South Africa also faces
constraints. Some of these goals, historically speaking, mean making difficult compromises and
trade-offs given current structures. Output growth affects emission growth and employment
growth. Employment growth affects social protection demand. Changes necessary to enable a just
transition of any interpretation require considerations of public debt levels as well as their impacts
on the balance of payments. If one of these goals is off-kilter, it has the potential to upset the
2 Besides for those with disabilities, who are offered support.
8
viability of achieving these goals together, with the failure to do so endangering long-term
sustainable development.
It is in this context that this research paper develops and implements a framework from which
improving outcomes in emission reduction, employment, and social protection coverage can be
more easily digested and considered, whilst considering what constraints are imposed in different
scenarios. What follows is a literature review in Section 2, and an analysis of historical trends and
contexts of emissions and energy, social protection coverage, and the labour market in Section 3.
Section 4 outlines the methodology employed in deconstructing the dynamics of the various goals
and constraints, with Section 5 harbouring their results and discussion thereof. Section 6 highlights
some of the main limitations of the paper, and opportunities for further work and development of
this analysis. Section 7 concludes.
2. Review of literature
Due to the wide scope and novel application of the techniques applied in this paper, a review of
literature is limited to a review of the gap in the literature, with some discussion on the papers’
motivations in terms of a strong sustainability approach.
Defining appropriate developmental trajectories is one of contention. What is generally agreed,
however, is that the concept of sustainable development outlined in the Brundtland Report (1987)
though contested in its specificities in practical application, is central to any vision. Perhaps best
illustrated by the Sustainable Development Goals (UN General Assembly, 2015), the principles of
sustainable development are many, and cut across the spheres of the environment, the social, and
the economic. The Brundtland Report (1987) further defines sustainable development as
“development that meets the needs of the present without compromising the ability of future
generations to meet their own needs”. However, due to the wide scope of the SDGs, there exist
synergies and tensions between them given historic trajectories and relationships between the
indicators measured. Whilst in many cases these relationships are symbiotic, for example between
clean water and sanitation (SDG 6) and good health and well-being (SDG 3), said synergistic
relationships lie mostly within non-environmental SDG targets. Pradhan et al. (2017) highlight
this in their analysis of SDG synergies and trade-offs, showing that, for example, SDG 12,
responsible consumption and production, is mostly associated with trade-offs with other SDG
targets.
What emerges is the need to recognise where these trade-offs exist, and how to manage and
mitigate the negative implications that arise with them. Wellbeing and the creation of wealth can
be viewed as the aggregation of four types of capital. These being manufactured (or produced),
social, human, and natural capital (Ekins et al., 2019). Traditionally, it has been the consumption
and inputs of natural capital that develop the other, mostly manufactured, types of capital.
9
However, as Ekins et al. (2019) succinctly point out is that it is unclear whether capital stock in its
totality needs to be constant or increasing for future generations to meet their needs. Further,
whether we can assume that like but opposite movements in different capital type levels result in
equal wellbeing levels is not clear. The appeal of sustainable development is near universal,
perhaps by construction of its vagueness – “development that lasts” (Atkinson et al., 1997) posits
an impossible-to-repost statement. However, greater specificity into the manner in which the four
types of capital interact with each other has been a subject of debate in the literature. Today, this
culminates in two concepts of sustainability: weak and strong.
In terms of wellbeing effects and wealth creation, the main contention between the concepts of
weak sustainability and strong sustainability is that of the appropriateness of the substitutability
between the types of capital involved. The weak sustainability approach, by it’s a priori that types
of capital are able to substitute each other in measurement of total wellbeing, precludes the obvious
notions of natural capital being distinct in its functions that are irreplaceable by other capital. Weak
sustainability, here, is key. It posits that it is possible for future generations to be adequately
compensated (Godin et al., 2022). However, this rests on assumptions that are obviously
unrealistic. For example, that the benefits derived from different elements of capital are uniform
in their substitutability i.e., that the consumption of natural capital can be offset by the increases
in manufactured capital. This implicitly involves being able to value capital in a common form,
currency, that has a range of issues for natural capital. For a more complete range of assumptions,
see Theys and Guimont (2019). The power and allure of weak sustainability comes from its relative
ease to implement in analysis due to it’s a priori of substitutive capacity of capital types. However,
this runs the risk of underestimating the damages to future generations due to the uncertainty
attached with environmental degradation and damage (Ekins et al., 2019). In weak sustainability’s
approach of the disregard for natural capitals specific characteristics, we undermine the value of
the environment as “a ‘glue value’ that holds everything together” (Turner et al., 1994), threatening
to drastically affect both the simply calculable capital stock that weak sustainability proports to
accomplish but also the efficacy of the other elements of capital to offset their growth’s effects.
The “weak” in weak sustainability is attributed by its detractors, those who promote a sustainable
development vision of its counterpart, strong sustainability. Strong sustainability’s primary
contention is with the substitutability of capital types inherent in the weak sustainability approach.
Here, there are proponents with a range of views regarding the degree of potential substitutability,
but the a priori that capital types are substitutable is dropped. This does not necessarily mean that
there is no potential for some degree of substitutability, but that the starting point of analysis is
that capital types are distinct (Godin et al., 2022). This makes room for more in-depth analyses,
often multidimensional, in assessing what more realistic trade-offs and synergies exist between the
capital types in a context of multiple developmental target goals. Where there are cases of
substitutability, there is room for a conclusion to be made inline with those of weak sustainability.
However, the a priori of strong sustainability allows for an analysis to identify where exceptions
10
exist, a crucial advantage. It is for this reason that the analysis to follow strives to employ an
approach of strong sustainability.
In terms of scope, an approach of strong sustainability requires many inputs. Indeed, Ekins et al.
(2019) highlight that the data requirements for one such measurement of environmental use, the
Environmental Sustainability Gap (ESGAP), has stringent data input requirements. In the context
of developing countries, this measure is not immediately implementable. Whilst imperfect, its
strength lies in attempts to providing a single indicator akin to that of GDP or the HDI metrics.
We know that for existing measurements, data measurement and availability have not yet garnered
the ideal traction, as evidenced for the HDI (Bonini, 2018).
The literature on implementing a Just Transition in South Africa is dense3 (World Resources
Institute, 2021), owing to its high unemployment rate and unique energy generation structure.
However, whilst numerous studies, roundtables, and policy papers have been developed in the
South African context, deeper methodological efforts at analysing development trajectories with a
nod to strong sustainability are sparse. In terms of environmental transitions, the literature is
focussed more on specific interactions. In South Africa, the main focuses are on employment
changes that come with alleviating coal dependence such as the work done by Hartley et al. (2019)
who show that changes from the 2016 to 2018 IRP would result in a net addition of 160 000 jobs
gained by 2050. Further, work by Hermanus and Montmasson-Clair (2021) supply the total
number of jobs at risk in the coal sector, whilst highlighting the lack of clear employment loss
mitigation plans in existence. Preliminary work by Pineda (2021) further shows how demand
shocks will impact employment by using input-output analysis, with net job gains from demand
shocks. By and large, focuses on employment changes are based on mitigation, not forecasting
specific results. Further analyses make inroads into identifying transition risks, such as those by
Godin and Hadji-Lazaro (2020) on financial fragility, Huxman et al. (2019) on risk allocation,
adaptation, financial stability, and sovereign ratings risk, and work on those most vulnerable to
employment-related transition risks (Makgetla et al., 2019). Naturally, these works come with a
range of policy proposals.
There are social considerations that come with green and just transitions too. In an environment of
high unemployment and inequality, this is a crucial consideration. In South Africa, there have been
various proposals to overhauling the social protection system in light of a gaping social safety net
for the working age population (Institute for Economic Justice, 2021; SAFSC, 2020; Senona, 2020;
Taylor, 2020). These proposals have been made in light of prevailing poverty and inequality,
though are relatively narrow in their scope. Climate destruction, however, reinforces and
intensifies the need for adequate social protection coverage (Costella et al., 2021; Rigolini, 2021)
through increased vulnerability, a shifting risk landscape, and mitigation benefits. Specific to
South Africa and a low-carbon transition, new work by Annecke and Wolpe (2022) has recognised
the importance of social and social protection policy in mitigating the harshest effects of a just
3 For an incomplete but important list of works, see https://lifeaftercoal.org.za/about/just-transition
11
transition – more than that of closing the existing gap in social protection coverage by means of
cash transfers. The authors further recognise the political intricacies of social policy analysis, an
important consideration in the South African context.
What these analyses are missing, however, is a unified framework in order to “bring it all together”.
Whilst important contributions exist, there is a gap in the literature on how to show, simply, how
these contentions fit together in a broader macroeconomic sense. What constraints exist from a
zoomed-out view has not yet been constructed on the whole, nor has the means to conduct such an
analysis been developed. The closest attempt in this regard is likely from (Espagne et al., 2021),
whereby an analysis of countries’ external, fiscal, and socioeconomic exposure to low-carbon
transitions are analysed. However, due to 189 counties being surveyed, it does not allow for the
adequate flexibility required for a more in-depth, country-level analysis, for obvious reasons.
Regardless, it is a powerful indicator of where structural vulnerabilities lie.
It is in this context that this paper is developed. Leveraging a preliminary framework by Godin
(2021) as a basis, we develop a methodology in order to assess the idiosyncrasies of the synergies
and trade-off for South Africa in the context of multiple goals in the environmental (emission
reduction), social (social protection coverage and adequacy), and economic (unemployment),
whilst considering the constraints that public debt and the balance of payments, along with
contemporary political economy, bring to the fore.
3. Historical Trends in Emissions, Social Protection, and
Employment
Contextualising the environment in which this research question operates in important. Below,
historical context is given to the main areas of interest in this paper: emissions and energy,
coverage and spending on social protection measures, and employment trends. By doing so, we
are able to situate the severity of the issues faced in these areas, as well as indicate what (if any)
improvements have been made over recent years.
3.1 Environment and emissions
The African continent is a small player in the global emissions sphere. We observe this in Figure
1, where only the Democratic Republic of Congo (DRC) and South Africa are represented in the
cohort of the 20 largest emitters of greenhouse gases (GHG) globally.
12
Figure 1. Total greenhouse gas emissions, 2018.
Source: Our World in Data (OWID) (2022).
Note: Filtered to show top 20 largest greenhouse gas emitters; those larger than 400 million tonnes of CO2e. Countries
are coloured by continent, with North and South America grouped together.
The emissions landscape is dominated by China, the US, and India. South Africa is, however, the
16th largest emitter of GHG globally in 2018, with some 1.06% of the global share. The DRC’s
high emissions are driven predominantly by land-use changes and forestry (LUCF), contributing
some 80.1% of their GHG emissions (ClimateLinks, 2018). With a closer look at Africa however,
we are able to see the extent to which South Africa contributes to GHG emissions, as illustrated
below in figure 2.
Figure 2. GHG emissions in Africa, 2018.
Source: OWID (2022).
Note: Filtered to show countries with 100 million tonnes or more.
13
As shown, the DRC and South Africa are by far the largest GHG emitters on the continent, with
Nigeria and Egypt also large emitters.4 Of further interest is the composition of GHG emissions.
To illustrate the specificity of South Africa’s carbon intensive GHG composition, figure 3
illustrates the share of CO2 in total GHG emissions. A more detailed breakdown of GHG emissions
by sector is available in the appendix.
Figure 3. Share of CO2 in GHG emissions, 1990 – 2018.
Source: OWID (2022).
Note: Largest four African GHG emitters included.
South Africa is by far the African country with the highest share of CO2 emissions contributing to
its GHG emission total. We even see figures of more than 100% between 1990 and 2000, and
again in 2009. We do however observe a slight decrease in the period 2010 to 2018, with this share
hovering between 87.7% and 96.6%. This is a clear reflection of the carbon-intensive nature of
South Africa’s GHG emissions. By contrast, the remaining three largest African GHG emitters
have much lower shares of CO2 in their GHG composition, with Egypt being the next highest,
Nigeria much less, and the DRC hardly having a CO2 contribution. Given the close link between
GHG emissions and CO2 emissions in South Africa, contextualising CO2-specific emissions is a
valuable exercise.
4 Though both with populations larger than South Africa.
14
Figure 4. Share of global CO2 emissions, 2019.
Source: Our World in Data (OWID) (2022).
Note: Colour of bar represents the continent of the country, with North and South America grouped. Africa is
represented by the peach shade.
Here, we see the CO2 emissions landscape dominated by China, the US, India, Russia, and Japan.
South Africa is however the 13th largest emitter of CO2 globally in 2019, with some 1.3% of the
global share. Within Africa however, the picture is much different to that of total GHG emissions.
Figure 5. African CO2 emissions, 2019.
Source: OWID (2022).
Note: Filtered to show countries with 15 million tonnes or more.
15
We observe that South Africa is by far the largest emitter of CO2 in Africa. Emitting some 476
million tonnes of CO2 equivalent (CO2e), more than twice the amount of CO2 than the next largest
African emitter, Egypt. Indeed, South Africa is responsible for the largest individual contribution
of CO2 emissions on the continent, though is a relatively minor player in the global context.
More specific to South Africa, it is worth noting the historical evolution of CO2 emissions. This is
illustrated by figure 6.
Figure 6. Annual CO2 emissions in South Africa, 1900 – 2019.
Source: OWID (2022).
Here we see the evolution of CO2 emissions in South Africa between 1900 and 2018. We see the
first rapid increase after the second world war until 1971, a three-fold increase in emissions from
55.9 MtCO2e in 1945 to 168.4 MtCO2e in 1971. We see this trend hiccup over the period from
1971 until the early 2000s, and increase rapidly over the period 2001 to 2009, peaking at 502.3
MtCO2e. Incidentally, this is also the period over which the commodity super cycle took place,
with mining output up over the same period. CO2 emissions have decreased and plateaued since
2010, not least due to the arrest of growth in energy generation from coal. This is not to say that
coal occupies a lesser share in energy transmission in South Africa, but that there has been an
arrest of growth in energy generation on the whole.
Turning to energy consumption, we see South Africa’s plateau in primary energy consumption in
figure 7. Here, we see an increase in annual consumption from 1965 to 2008 of 354 terawatt-hours
to 1458 terawatt-hours, a huge increase over the period. Over the following decade however,
energy consumption plateaued, measuring just 12 terawatt-hours higher in 2018 than in 2008, with
a more recent decrease in 2020 to levels below those in the late 2000s.
16
Figure 7. Primary energy consumption by source, 1965 – 2020.
Source: OWID (2022) & BP Statistical Review of World Energy (2021).
Further, we see the composition of energy by source. Coal is by far the largest source of energy,
followed by oil (British Petroleum, 2021). It is here that the CO2-intensive GHG emission
landscape of South Africa becomes clear. Alternative sources of energy (solar, wind, hydro,
nuclear, gas, and other renewables) comprise a small fraction of energy consumption. This is
crucial, as energy generation contributes a large portion of CO2 and hence GHG emissions in South
Africa.
It is likely no coincidence that this plateau in energy consumption has occurred over the same
period that South Africa’s “loadshedding” phenomenon has become a key feature to energy
generation and use management. Due to demand for energy outstripping its available supply, ailing
state-owned energy producer Eskom restricts energy use on a rotational basis, with a variety of
severities (frequency and scope of areas affected). We can take two key insights from a zoomed-
out view. First, that South Africa’s energy generation system cannot support the needs of South
African people. Breakdowns, shortages of materials, and maintenance in a single energy
generation plant can cause the need for loadshedding/load reduction mechanisms nationwide. It is
clear that the system is fragile. Secondly, is that there has been little capacity added to the energy
generation system in a decade. Underinvestment, the precarious financial position of Eskom, and
mismanagement (including corruption) have put the energy security of South Africa under threat.
17
The effects of this are more visible in figure 8 below, which show the effects of the lack of growth
in energy consumption on the average South African.5
Figure 8. Primary energy consumption per capita, 1965 – 2020.
Source: OWID.
Here, we see that average primary energy consumption since 2008 has been on a sharp downward
trend. Primary energy consumption per person has seen an 18.6% decrease from 2008 to 2019. We
may see this as a reversion of the trend seen from 2000 to 2008, over which South Africa
experienced a much higher economic output growth rate relative to the following decade. As
corroborated here, energy consumption is positively correlated with economic growth. From 2008
onwards, it has been such that modest population growth, compounded, combined with decreasing
energy use, has led to these decreases per capita. 2020 sees a continuation of the downward trend,
though this is in the context of COVID-19-related phenomena, and not necessarily reflective of
medium-term energy consumption per capita trends.
Considering this, it is worth turning to South Africa’s energy generation efficiency. By this we
mean to assess whether the efficiency of energy generation has improved or worsened over the last
three decades.
5 This is simply total primary energy consumption divided by South Africa’s population, not indicative of actual
energy use of any single person.
18
Figure 9. Intensity of CO2 emissions vs energy generation, 1990 – 2019.
Source: Authors construction. OWID (2022), International Energy Agency (IEA) (2021), Enerdata (2021), World
Development Indicators (WDI) (2022).
Note: CO2 measured in millions of tonnes, energy in petajoules.
Figure 9 highlights the evolution of South Africa’s CO2 intensity with respect to energy generation.
There was a marked increase in intensity between 1990 and 2004, up from 0.145 Mt per PJ in 1990
to just shy of 0.18. There was a plateau around this level until 2010, whereafter there was a sharp
decline to the 0.16 level by 2013, where the intensity has hovered until its last measurement in
2019. We may view this as a period wherein higher economic growth during the commodity
supercycle pushed intensity upwards, leaving the intensity to subside after the GFC, with CO2
intensity reverting to levels similar to those of 1995 – 2002.
In terms of how useful this is in generating output, we can assess exergy: the efficiency of energy
available to do work. In this case, we assess exergy with the input being available energy
generated, with the work corresponding, in macroeconomic terms, to output (GDP). Figure 10
creates a ratio of energy use vs GDP, creating a useful visualisation of South African energy-use
efficiency. We observe, on the surface, a vast improvement of energy use from 1990 to the most
recent measurement. From a 1990 figure of over 0.9 petajoules (PJ) of energy use per R1 billion
of output, this has improved to 0.6PJ per billion Rand on output in 2019. This is a positive
movement in terms of energy-use efficiency. However, we can see that South Africa is likely
reaching a plateau within the current technological landscape, with no meaningful improvements
since 2010.
19
Figure 10. Energy use per R1 billion GDP, 1990 – 2019.
Source: Authors construction. OWID (2022), IEA (2021), Enerdata (2021), WDI (2022).
Note: Exergy is PJ energy per R1b GDP (constant).
Clearly, South Africa faces a need to reduce its carbon footprint. As the highest emitter of CO2 in
Africa, among the highest in the world, the transition to a more environmentally friendly dynamic
of production and consumption will be difficult. Driving this difficulty is the phenomenon of the
mineral-energy-complex (MEC). The MEC creates issues in two manners. Firstly, it is tied
inexorably to the process of energy generation of South Africa. Shares of coal-use in the South
African energy generation system vary, with the IEA (2021) citing some 72%, and the Department
of Mineral Resources and Energy (DMRE) citing 77% (2022). In either case, South Africa’s
energy generation structure is dominated by coal. By contrast, renewables accounted for just 7.6%
of energy generation in 2020, up from 2.1% in 2010 (Enerdata, 2021).
It is in this context that South Africa finds itself in terms of reducing CO2 emissions. Doing so
necessitates a move away from coal dependence for its energy needs, as highlighted by the recent
announcement made at the Conference of the Parties (COP26). South Africa entered into a
preliminary commitment with the United Kingdom (UK), United States, France, Germany, and
the European Union (EU) for a facility of US$8.5 billion to achieve this (COP26, 2021). The
conditions of this agreement are still unclear, putting a damper on the viability of the funding.
Details available regarding the scope of the agreement include simply “appropriate financial
instruments, which may include but is not limited to multilateral and bilateral grants, concessional
loans, guarantees and private investments, and technical support to enable the just transition, with
a view to longer-term engagement” (COP26, 2021). It must be noted that multilateral funding,
even if concessional, has not been a positive feature of countries of the Global South. The
conditionalities of this agreement are important, and until those details are clear, it is difficult to
assess how useful this funding will be and what it will be used for. Further, Eskom is embattled
with a crippling debt bill, some US$26 billion, and having received at least US$8 billion in bailouts
over the last 12 years (Faku, 2020). Whether the energy giant has the capacity to implement this
funding properly is yet to be seen. In a broader scope, financing requirements on a true green
transition are scarce, though Eskom itself tentatively estimates that it would require a figure in the
20
area of US$27 billion (Burkhardt, 2021). It is therefore no small portion of this requirement that
the JET Agreement covers: 31% of total required financing.
South Africa’s future energy generation plans hinge mainly on the Integrated Resource Plan (IRP)
of 2019. The IRP mainly focuses on coal-fired expansion (Department of Energy, 2019). The plan
is clear about the main feature of energy generation being through coal, though stating “Carbon
capture and storage, underground coal gasification, and other clean coal technologies are critical
considerations that will enable us to continue using our coal resources in an environmentally
responsible way into the future” (Department of Energy, 2019). This is under the heading of
“Environmental Considerations”, which is hardly half a page in length. It is thus clear that there is
a motive to continue along the path of commoditising coal, particularly in terms of carbon trading
and sequestration measures, as has been the case elsewhere (Spash, 2014). More revealing is that
by 2019, only one plant complied with nationally legislation on air pollution standards after Eskom
applied for a 5-year extension on this compliance in 2014. Whilst the greenwashing of coal in the
IRP 2019 is clear, DMRE Minister Gwede Mantashe has taken to other means of public persuasion
against ending the dependence of coal. In a speech to the African Energy Week conference,
Mantashe touted that “I think Africa must get together to develop a strategy to deal with this reality.
Africa must seize the moment, we must indeed position Africa oil and gas at the forefront of global
energy growth” (Steyn, 2021), citing coercion of Africa to take developmental missteps by an anti-
fossil fuel agenda. This is in contrast with President Cyril Ramaphosa’s words in a statement “The
time for greater climate action is now. We have to reduce our emissions. We have to adapt and
build resilience for our communities and for our economy” (Ramaphosa, 2021). With Mantashe at
the helm of long-term energy generation planning, it is difficult to see South Africa meeting its
Nationally Determined Contributions (NDCs) by 2025 or 2030. Liquified natural gas (LNG) was
more recently championed as a bridging solution away from South Africa’s coal dependence,
though with recent increases in LNG prices as well as a process of bidding that is not well
supported and politically uneasy, this is unlikely to happen in the short term (Comrie, 2022).
In order for South Africa to meet its Nationally Determined Contributions of between 398 and 510
MtCO2e in 2025 and 350 to 420 MtCO2e in 2030, reductions of 17% and 32% are required by
2025 and 2030 respectively (NDC Partnership, 2021). These low and high ends of these ranges
correspond to the 1.5 degree pathway and 2 degree pathways respectively.
21
Figure 11. Indicative reduction of CO2, 2016 vs 2021 NDC update.
Source: Tyler and Steyn (2021)
Note: 2017 emissions are from last measurement at the time of this report release (CO2). In 2019 it was 475Mt.
South Africa is already within its 2025 upper-bound target range. However, with less than 4 years
to achieve the first goal on the lower-bound end, no meaningful reductions in emissions have been
seen thus far. What the updated September 2021 NDC targets have shown us, however, is that a
more ambitious climate-friendly development pathway has been envisioned (compared to the 2016
NDCs). For these ambitious targets to be met, however, accompanying policy mechanisms and
action is required alongside the document that declares the targets. The NDC Partnership (2021)
further estimates that achieving this will require financing by the international community of US$8
billion annually if this is to be met alongside a just transition. Considering the lack of details and
urgency of the COP26 US$8.5 billion financing commitment to be disbursed over the next 3 to 5
years, the situation seems unlikely. Further, frictions in the labour market and output that may
accompany ill-developed industrial policy may lead to worsening outcomes in the social sphere.
3.2. Social protection coverage
The social safety net in South Africa has grown in scope drastically since the end of Apartheid in
1994. This has been due to the introduction of new mechanisms to protect the most vulnerable in
society, as well as those previously deliberately and arbitrarily excluded from adequate support
now being able to access it. Here, we consider specific social assistance coverage and adequacy,
as a subset of broader social protection measures (Farrington and Slater, 2006). In particular, we
focus on the gap in the South African social protection net for the working age population (18 to
59 years of age), as well as the adequacy of the Child Support Grant (CSG) targeted to children
between 0 and 18 years of age.
Whilst cash transfer mechanisms of social assistance and social protection are analysed here, it is
important to consider that social protection takes many other various forms. For example, whilst a
22
proposal of a Universal Basic Income Guarantee (UBIG) is made below, an argument for the
improved provision of basic services is as important. The breadth and quality of healthcare,
education, and infrastructure provision are crucial (Goldman et al., 2020), and should be seen as
complementary to direct social assistance measures. Indeed, it is not the case that increased
spending on grants is a panacea to increasing welfare and poverty dilemmas. There is no “silver
bullet” in this regard. However, cash transfers play a large role in the South African social
protection schema, and have shown impressive, positive impacts on recipients, their households,
and their surrounding communities. Further, a focus on cash transfers allows fiscal expenditure
estimates and forecasts to be made with greater accuracy due to the more basic calculation
procedure as well as the availability of input data required. It is for these reasons that the below
measures are considered in isolation, with additional measures being outside the scope of this
analysis.
At first glance, it may seem as though South Africa has increased social assistance spending
drastically since 1994. However, it is important to contextualise this in light of South Africa’s
racially exclusionary past. As a brief indicator of the extent to which this is relevant, figures 12
and 13 below highlights some of the misalignment between need and service in the 20th century.
Figure 12. Average annual value of State Maintenance Grants, 1990 rands.
Source: Lund Committee (1997).
Here we observe the differing annual amounts paid for the same grant, with white recipients
receiving far greater sums per annum than other racial groups. Black recipients in particular
received much lower annual sums, some 20% of the amount disbursed to white recipients in 1975,
for example. There is a convergence towards 1990, however, though some non-trivial disparities
still existed between the racial groups.
23
Figure 13. Number of State Maintenance Grants per thousand children.
Source: Lund Committee (1997)
Further concerning, and more illuminating on the issue of exclusion from social security measures,
is figure 13. Here we observe the disparity between racial groups receiving the State Maintenance
Grant (SMG) between 1962 and 1991. Coloured and Indian recipients were far ahead of White
and Black recipients. However, relative to general standards of living between the racial groups,
the figure of 2 recipients per thousand children in 1991 for Black recipients is indicative of the
system’s blatant discrimination. Those in greatest need of support were not able to access and
benefit from state support. More overtly racist overtones existed prior to this too, by example of
Black and Indian people being excluded from receiving old-age pensions on the basis that
customary extended familial support would suffice as a community safety net (Reddy and
Sokomani, 2008). As can be seen above, the level of grants disbursement was also reduced for
white recipients over time. This is in response to the stress that including different racial groups
more actively in the social security scheme put on the systems fiscal viability (Van der Berg, 1994).
The threat of fiscal viability is still used in contemporary discourse surrounding the expansion of
social assistance measures, though with the explicitly racist overtones being supplemented for
those of classist disdain.
Today, South Africa’s social security net is wide-reaching. From some 3 million beneficiaries in
1995, South Africa today services approximately 18.5 million beneficiaries on a monthly basis
(Ramaphosa, 2022). Below, the most recent figures of recipients and their individual disbursed
amounts are provided.
24
Table 1. Overview of social grant landscape in South Africa.
Grant Type
Number Recipients
Amount 20226 (ZAR)
Approx. Annual Cost (bn ZAR)
Care Dependent Grant
151 366
1 990
3.615
Child Support Grant
13 038 890
480
75.104
Foster Child Grant
340 401
1 070
4.371
Disability Grant
1 043 691
1 990
24.923
Grant in Aid
271 960
480
1.566
Old Age Grant
3 713 198
1 990
88.671
War Veterans Grant
747 720
2 010
0.747
Total
18 559 537
198.251
Source: Authors construction. South African Social Security Agency (2021b)
Note: Figures used are most recent available, corresponding to end of August 2021. There is some variation each
month, though none material in the short term.
The Child Support Grant (CSG) and Old Age Grant (OAG) are by far the widest reaching, with
13 and 3.7 million beneficiaries respectively. Of note here is the respective expenditure on these
two items. The CSG provides for three and a half times more South Africans than the OAG, though
only sees 85% of the latter’s expenditure. There is an argument to be made surrounding the level
of the CSG too, considering the levels of the National Poverty Lines (NPL). The NPLs are
determined by the national statistical agency, Statistics South Africa, and correspond to different
levels of needs being met.7 Three poverty lines exist, with definitions supplied by Statistics South
Africa (2021b):
• Food poverty line (FPL), R624/month: “the amount of money that an individual will need
to afford the minimum required daily energy intake. This is also commonly referred to as
the “extreme” poverty line.”
• Lower-bound poverty line (LBPL), R890/month: “the food poverty line plus the average
amount derived from non-food items of households whose total expenditure is equal to the
food poverty line.”
• Upper-bound poverty line (UBPL), R1 335/month: “the food poverty line plus the average
amount derived from non-food items of households whose food expenditure is equal to the
food poverty line.”8
Clearly, the amount that the CSG is set at is well below that of the FPL. That is, that the grant
cannot meet the minimum required daily energy intake for one person. Further, the grant is well
below the FPL, requiring a 30% increase to be on par. Increases announced for FY2022 constitute
R20, or a 4.3% increase (Human, 2022). This is in the context of inflation of 5.7% for 2021.
Notwithstanding the initial inadequacy of the CSG level and the havoc brought about to the South
African economy and its social setting by COVID-19, state support for the most vulnerable
children is being withdrawn in real terms. With 65% of children aged 0 to 17 receiving the CSG
6 For FY2022, grants increase in April and again in October. The figures shown are the planned October 2022
amounts.
7 Constructed through the use of the 2010/11 Income and Expenditure Survey’s (IES) household expenditure data,
which likely does not still accurately reflect consumption patterns and composition a decade later.
8 Italics provided by the author for ease of differentiation.
25
(South African Social Security Agency, 2021b), this is a major concern for one of the major
developmental tools in South Africa.
South Africa does, however, spend a large portion of their total consolidated expenditure on social
protection measures. Figure 14 below highlights this, as well as the planned expenditures on social
protection into FY 2024/25. We observe an increase in the ratio of social protection spending vs
total spending in 2020, unsurprising due to the necessity that COVID brought. However, this
decreases in 2021 and remains flat in 2022, before a planned decrease in relative terms in 2023
and beyond. FY 2024/25 sees this ratio at 15.02%, lower than any year of observation since FY
2005/06, a sharp reversal of support. This is also measured against a decreasing rate of total
expenditure growth. Clearly, the South African National Treasury envisages less demand for state
support. How this view is justified is unclear.
Figure 14. Total and social protection expenditure, 2005/06 – 2024/25.
Source: National Treasury (2022).
Note: SP refers to social protection, which includes measures in addition to social grant expenditure. Left axis is
spending, right axis is ratio in %.
It is worth considering what affect a potential increase in the CSG to the FPL would be in this
context. Considering a constant rate of 65% of children between the ages of 0 and 17 receiving the
grant each month at an increased FPL amount of R635, increasing incline with population
estimates and forecasted inflation,9 total expenditures are as follows:
9 As per National Treasury forecasts until 2024, with 4.5% thereafter.
0,00
5,00
10,00
15,00
20,00
25,00
-
500,0
1 000,0
1 500,0
2 000,0
2 500,0
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
SP/Total (%)
Spending (ZAR bn)
SP Spending Total Spending Ratio
26
Table 2. CSG spending scenarios in the medium term.
2022
2023
2024
2025
Total Population
20 310 752
20 421 170
20 497 978
20 541 755
Total Receiving
13 201 989 13 273 761 13 323 686 13 352 140
CSG + 0 (ZAR)
480
503
526
551
CSG @ FPL (ZAR)
624 654 683 713
Total + 0 (ZAR bn)
76,04 80,13 84,05 88,27
Total @ FPL (ZAR bn)
98,86
104,16
109,16
114,31
Additional Spending (ZAR bn)
22,81 24,04 25,11 26,04
Source: Authors construction using UN Population Estimates (2022), South African Social Security Agency (2021b),
National Treasury (2022b).
Note: Increases are estimated at forecasted inflation. Total population refers to those between the ages of 0 and 17
inclusive. CSG + 0 refers to current CSG levels without increases. CSG @ FPL refers to increasing the CSG to the
level of R635/month (the FPL). Total figures are in ZAR, corresponding to annual expenditure of the two amount
regimes. Additional spending refers to the different between the current CSG level and that of the CSG at the increased
FPL amount.
Evidently, the additional expenditure of increasing the CSG level to that of the food poverty line
is not insurmountable on an annual basis. This increase would likely see improved outcomes for
young recipients as well as those in their households too, whilst being in line with constitutional
obligations (see below).
To continue with the coverage of the vulnerable with appropriate social protection measures, it is
worth a brief analysis of those uncovered by existing measures. As indicated, the main two cash
transfer mechanisms cover those aged 0 to 17 years of age (the CSG) and those over the age of 60
(the OAG). However, no measure broad measure exists to cover those of working age (18 to 59).
What does exist are measures that cover those living with disabilities, as well as a labour market-
linked contribution-based national unemployment insurance scheme.10
This is particularly relevant to the context of the state’s constitutional obligations. The South
African Constitution, Section 27 (Republic of South Africa, 1997) reads: “
(1) Everyone has the right to have access to –
a. Health care services, including reproductive health care;
b. Sufficient food and water; and
c. Social security, including, if they are unable to support themselves and their
dependants, appropriate social assistance.
(2) The state must take reasonable legislative and other measures, within its available
resources, to achieve the progressive realisation of each of these rights.”11
With high levels of poverty and lack of access to income earning opportunities, there is a clear
contradiction between the above constitutional mandate and the lack of appropriate social
10 The UIF, the Unemployment Insurance Fund.
11 Italics added for emphasis by the author.
27
assistance for the working-age population. Addressing this gap is in terms of coverage and
adequacy is therefore not an unfounded plea; it is a call to action for the government to fulfil its
obligations.
More broadly, the International Labour Organisation (ILO) provides us with an overview of social
protection coverage, as indicated in table 3.
Table 3. Social protection coverage in South Africa, 2020.
Source: ILO (2022)
COVID-19 brought with it the first instance of broader coverage for the working age population
with the introduction of the Social Relief of Distress Grant (SRD). Introduced in April 2020 (The
Presidency, 2020), the grant aimed at providing support of R350 (US$23)12 per month to those
with no other income in a time where lockdown restrictions on movement and workplace
interaction was high (University of Oxford, 2022). In light of South Africa’s nexus of issues
involving poverty, inequality, and unemployment, and with the labour market failing to provide
an avenue of poverty reduction from at least the GFC of 2008, this was an incredibly necessary
and important social policy.
Despite its drawbacks, especially in the early stages of implementation, the SRD grant has been
crucial to the social wellbeing of the most vulnerable in South African society. Faced with a
supposed precarious fiscal position, functionaries within government (mainly National Treasury)
have tried to end the support measure. This has faced strong pushback from the public, civil
society, and academics, ultimately prolonging its implementation from the initial 6-month period
to what was recently announced to be until December of 2022. Additionally, there have been calls
to transform the SRD grant into a Universal Basic Income Guarantee (UBIG) by civil society,
trade unions, and academics, a social policy debate now revived since its first serious consideration
in the late 1990s/early 2000s.
Several papers detail the positive impacts of the SRD grant during its implementation over 2020
and 2021.13 The primary socioeconomic panel dataset in South Africa, the National Income
Dynamics Study (NIDS), was discontinued in 2018 due to lack of funding, though found renewed
support from private donors to assess the socioeconomic impacts of COVID-19, though through a
reduced sample and in the form of the National Income Dynamics Study-Coronavirus Rapid
Mobile Survey (NIDS-CRAM).
12 At an exchange rate of R15/US$
13 Which are available at https://cramsurvey.org/reports/
28
Primarily, the findings show that the grant gave support to those in need. Recipients were those in
greatest distress and precarity along the income distribution (Jain et al., 2020), as well as areas that
experience the highest rates of poverty (Department of Social Development, 2021). Further, the
grant mitigated household hunger measures by significant margins, as well as decreased income
inequality by 3 percentage points14 (Republic of South Africa, 2021a). Further, grant recipients
reported that the vast majority of the grant amount was used to purchase food items that otherwise
would not have been affordable. Additionally, 70% of recipients we below the age of 34, providing
support to younger people with relatively smaller adaptive capacity to economic shocks. Clearly,
the grant was effective in its aims to provide support to those in need. Further, those investigating
the impacts give recommendations to government that detail the need to extend its implementation
(Stent, 2021; van Seventer et al., 2021).
However, there are important limitations to the impact of the SRD grant.15 Firstly, the level at
which the grant is set is less than 60% of the FPL (Statistics South Africa, 2021b). As such, those
without any other source of income are unable to survive on the grant alone. In practice, those
receiving the SRD grant are often in households that also receive some other social assistance,
commonly in the form of the CSG or OAG. This means that, as was common before the pandemic,
funds are pooled and distributed amongst members in the household, diluting the amount allocated
to the intended recipients (children and the aged), with negative social welfare effects in relative
terms for those recipients. The amount of the grant is too low. It is less than a dollar a day and
impossible to survive on alone, reducing the potential power of the grant to increase social welfare.
Second, the qualifying criteria necessary to receive the SRD grant was initially incredibly
exclusionary. This occurred in two ways. One, that the criteria of application were inappropriate,
with restrictions including the receipt of no money into one’s account at all, regardless of the
source. Remittances or familial support transactions for survival for the poor therefor disqualified
receipt. Further, only South African residents were permitted to receive the grant, excluding a
cohort of migrants who are already in a precarious environment and face overt political and social
xenophobia. Applicants were also not to be receiving any other grant, and not be supported by
government in any other way (student funding, insurance benefits, not a resident in a government
funded or subsidised institution) (Department of Social Development, 2020). Further, the
documents required for verification were not appropriate. Here applicants were required included
those which many South Africans do not have, and those even less common to be in the possession
of the most vulnerable such as banking details, proof of residential addresses, and cell phone
details. This leads one to believe that unnecessarily technocratic avenues were being used to verify
the needs of those in precarious situations, irrespective of the well-known context and environment
in which these people live. In addition to these inefficiencies, the manner in which applications
were verified was ineffective, reflected in the number of undue rejections of grant applications
14 This is reported by the government and Minister of Social Development, though without underlying support.
15 For more details, see https://www.timeslive.co.za/news/south-africa/2021-07-27-r350-social-relief-grant-mired-
in-challenges-despite-its-benefits-says-black-sash/ and https://c19peoplescoalition.org.za/
29
(C19 People’s Coalition, 2021). Here, outdated records of tax returns, bank account data, and
unemployment insurance data were used to qualify applicants. Worthy of mention is the gendered
nature of exclusion, with women only representing 37% of accepted applicants, although facing
harsher burdens of care and underrepresentation in the labour market. Due to sustained pressure
by civil society and public outcry of unfair exclusion, this was investigated, and the dataset used
determined to be unsuitable.
Regardless, the SRD grant has been a lifeline to some 10.5 million of South Africa’s most
vulnerable people. There have been calls to extend this measure permanently, with the most
suitable version being dubbed a Universal Basic Income Guarantee (UBIG) (Institute for
Economic Justice, 2021).16 In assessing the cost implications of a UBIG, a similar typology is used
to that of the Institute for Economic Justice (2021). All those aged 18 to 59 are considered, with
levels of 80% and 60% corresponding to eventual and initial uptake rates respectively.17 Those
employed in the informal sector are considered, those unemployed by the broad definition, those
not economically active, and finally all those not employed in the formal sector (NFE; not formally
employed). Those NFE are most likely to self-select for potential UBIG receipt, and this is where
the upper end of cost estimates should be limited.
Table 4. Annual cost of a UBIG (ZAR bn), 2022.
Group (18 - 59)
Number of people
FPL (R624 pm)
LBPL (R890 pm)
UBPL (R1335 pm)
All
34 724 361
260,02
370,86
556,28
All (80%)
27 779 489
208,01
296,68
445,03
All (60%)
20 834 617
156,01
222,51
333,77
Informal Sector Workers
2 749 846
20,59
29,37
44,05
Unemployed (b)
11 673 831
87,41
124,68
187,01
Not Economically Active
12 665 304
94,84
135,27
202,90
NFE
24 543 455
183,78
262,12
393,19
NFE 70%
17 180 419
128,65
183,49
275,23
Source: Authors construction using Quarterly Labour Force Survey (QLFS) Q4 2021 via DataFirst (2022).
Note: Includes agricultural workers. Those unemployed (b) refers to those both unemployed and discouraged workers;
the broad definition. NFE refers to those not formally employed; those not in the formal sector. Those employed in
private households are not considered to be in formal employment.
As evident, a UBIG set at the FPL of R624/month would require an additional R184 billion in
expenditure on social protection. Even this figure should be viewed as an unlikely upper-bound
estimate. If for example, there is only a 60-80% uptake of this group, this falls to an expenditure
of R110 to R147 billion. Evidence from the SRD grant have shown that 10.5m people have
benefitted, despite some 16m people being eligible, giving credibility to the lowering of receipt
expectation. As such, we make use of a midpoint value of 70% of those NFE as a most realistic
estimate of uptake, corresponding to R129 billion additional expenditure in 2022 terms. Even so,
16 A policy brief developed as main author.
17 Based on uptake rate experience of other grants.
30
this would be an upper-range estimate based on the 10.5 million uptake the SRD grant. There have
been numerous budget-neutral financing mechanisms made to fund this, despite the large value.
With respect to the final additional expenditure of the increase in value of the CSG and the
introduction of a UBIG create, social protection expenditure is illustrated below:18
Table 5. Additional spending for increased social protection coverage (ZAR bn), 2022-2025.
2022 2023 2024 2025
CSG @ FPL 22,81 24,04 25,19 26,38
UBIG @ FPL, 70% NFE 128,65 134,82 140,75 147,09
Total additional 151,46 158,86 165,94 173,47
Source: Authors construction.
With this in mind, it is important to consider a harsher relationship: demand for adequate social
protection and its affects on the fiscus are directly related to the dynamism and absorptive capacity
of the labour market. The support level considered above is meagre by any measure: just enough
to ensure sufficient monthly caloric intake. It is the number of those in need that is a driving force
of the total additional financing required. 65% of children live in households under the means-
tested threshold for the CSG. Unemployment is at record highs, with only 9.28 million South
Africans in formal employment, an expanded unemployment rate of 44.6%. This support is crucial,
though its necessity an indictment on the labour market as a tool of poverty alleviation, as well as
the lack of functioning industrial policy as a vehicle for meaningful employment growth and
development.
3.3. Employment and labour
South Africa faces an acute unemployment crisis. A system of deliberate educational and labour
market segregation and exclusion until 1994 left the South African government a mammoth task
of restructuring the productive forces of the South African economy. 1994 saw some 8.9 million
people employed with a corresponding 20% unemployment rate. By 2019, South Africa almost
doubled the number of those employed, with 16.4 million people employed in the last quarter of
2019 (Statistics South Africa, 2022a), though with unemployment reaching 29%. This evolution
of employment and unemployment has not been a linear, however. Figure 15 below highlights
this.
18 How this affects the proportion of social protection spending vs total consolidated expenditure can be found in
the appendix.
31
Figure 15. Total employment, 2000 – 2022.
Source: Authors construction, Statistics South Africa (2022).
Here we see two main periods of employment gains, and two employment crashes. Employment
gain periods are from 2004 to 2008, and from 2011 to 2020. Sharp decreases in employment are
seen in 2009 and in 2020, corresponding to the great financial crisis (GFC) and COVID-19
pandemic. Given these incidences of employment loss, 7.6% from peak to trough from the GFC,
and 13.8% from COVID-19, we see the responsiveness to the labour market to external and local
stimuli. The COVID-19 shock, in particular, is severe, removing all employment gains made since
2013. Regardless, there has been modest employment growth in absolute terms; an upward trend.
Figure 16. Employment growth rate, 2000 – 2022.
Source: Authors construction, Statistics South Africa (2022).
Note: Black line is quarterly growth, red line is 3-quarter smoothing.
A closer look at employment growth rates shows us two distinct phases of employment growth.
Here, it is more relevant to show relative rates. We observe strong though erratic employment
growth pre-GFC, with a less volatile but lower rate thereafter. This low employment growth rate,
32
post-GFC, has been persistent for the better part of a decade, despite the various initiatives to
address the employment crisis in South Africa. We also observe the impact of the COVID-19
pandemic, illustrative of the devastating effect on employment.
We are also able to observe where this growth in employment has come from by looking at sectoral
employment trends.
Figure 17. Sectoral employment by total share, 2000 – 2022.
Source: Authors construction, Statistics South Africa (2022).
Note: Filtered by final size of employment share in Q4 2021.
We see the majority of employment growth being in already large sectors, specifically in
community services and finance. Construction saw impressive growth since 2000, though was
hard-hit by COVID-19. Wholesale and retail trade saw a plateau in employment since 2010, with
a decrease due to COVID. Transport saw steady gains, though marginal in the larger picture.
Interestingly, we see a large decline of some 406 thousand jobs in the manufacturing sector
between Q4 2008 and Q1 2020, and another 389 thousand since then, effectively decreasing total
employment in the sector since its 2008 peak by 38%. In a broader sense, we see a shift towards
employment in the services (tertiary) sector of the economy, with both primary and secondary
sector employment shares shrinking.
33
Figure 18. Total employment share by sector, 1993 – 2019.
Source: Strauss et al. (2020).
Unfortunately, employment growth has been too modest. Relatively growth in the labour force has
been much larger, and unemployment has worsened.
Figure 19. Narrow unemployment rate, 2000 – 2022.
Source: Authors construction, Statistics South Africa (2022).
Note: Black line is actual, red line trend uses 3-quarter smoothing.
As illustrated, unemployment has, despite progress made over the 2000s, not been able to
meaningfully decrease. The period 2003 to 2009 saw improvements to the rate, with large
employment growth relative to labour force growth, carrying a level between 21% and 24% from
34
2005 to 2009. 2009 proved a turning point, however. A rapid increase between 2009 and 2010,
followed by a holding of the unemployment rate between 2010 and 2016, devolved into a climbing
unemployment rate between 2016 and Q1 2020, where the rate stood at 30.1%. The impact of
COVID-19 saw a large reduction in the unemployment rate, though only due to the shrinking of
the labour force due to restrictions and regulations implemented during the most stringent stages
of the south African containment response. With the return of the more usual labour force
contingency, unemployment increased drastically, above 32%, to the 35.3% at last measurement.
Measures implemented by the state to stem job losses were relatively strong, though the economy
still bled some 1.84 million jobs.19
Comparing this to targets of the National Development Plan 2030 of 2013, employment has
arguably been the most difficult target to meet. We see this below, with a comparison of the targets
versus the actual unemployment rate.
Figure 20. NDP unemployment targets vs reality.
Source: Authors construction. Stats SA (2022) & National Planning Commission (2013).
Note: Quarterly average unemployment rate used for reality of unemployment figures shown.
Evidently, the plan has not come to fruition. Targeted and actual unemployment rates have moved
in opposite directions, with the gap between the unemployment rate in 2020 being some 15%
higher than the envisioned target. At the end of 2021, this figure is even more dire, with
unemployment at a record high of 35.3%. Despite the various iterations of industrial policy since
1994, the trend in unemployment since 2015 has been detrimental. More concerning is the lack of
response from the National Planning Commission, the state at large, and the private sector. High
unemployment, it seems, will continue to be a feature of the South African economy in the absence
of radical restructuring of productive and (re)distributive processes.
Difficulties in generating sufficient income generating opportunities for a greater part of the South
African population has stunted output growth as well. We see this below, where there is a stalling
of output growth, especially evident from 2013 (post-GFC rebound).
19 Q1 2020 (pre-pandemic) vs Q4 2021 (latest).
25
20
14
6
24,9 25,3
29,2
0
5
10
15
20
25
30
35
2010 2015 2020 2030
Unemployment rate (%)
NDP Target Reality
35
Figure 21. Constant GDP, 2000 – 2021.
Source: Statistics South Africa (2022b).
Note: Measured in constant 2015 ZAR.
Here we observe two distinct epochs of output growth: from 2000 to 2008, and from 2009 to 2019.
The former period saw strong output growth, an increase of 38.6% over the 8-year window,
followed by a weaker period of just 18.6% growth over the next 10 years. COVID-19 may bring
about another change in growth prospects, different from the lacklustre preceding decade. Whether
this will be in the direction of a sustained period of stronger or weaker growth, however, is
uncertain.
Visually, we can see the changes in output growth in figure 22.
Figure 22. GDP growth, 2001 – 2021.
Source: Authors construction, Statistics South Africa (2022b).
Note: Output measured in constant 2015 ZAR.
Similar to employment, we observe a similar pattern of growth in productivity pre-GFC, with a
plateau and decrease thereafter, with productivity depressed from 2014 to 2020. 2021 sees an
uptick in productivity, though this is largely due to output recovering with greater impetus
compared to employment. This is illustrated below in figure 23.
36
Figure 23. Productivity, 2000 – 2021.
Source: Authors construction, Statistics South Africa (2022a, 2022b).
Note: Productivity is calculated as total output in constant 2015 ZAR divided by total employment.
Figure 24 allows for a clearer analysis of productivity growth over the last two decades too, with
strong growth in the early 2000s, negative growth between 2012 and 2017, and upticks over the
recoveries from the GFC and COVID-19 pandemic crises. Importantly, there is no substantial
movement in productivity since 2010. In fact, the slight decrease since then (pre-COVID), is
indicative of stagnant technology capacity. A decrease in productivity could, in the short term, be
seen in a positive light too. This would be the case of increasing employment diluting output,
where there would be an expected output response in the future.20 However, even this failed to
materialise, as unemployment continued to rise.
Figure 24. Growth in productivity, 2001 – 2021.
Source: Statistics South Africa (2022a, 2022b).
Note: Productivity is calculated as total output in constant 2015 ZAR divided by total employment.
20 Notwithstanding the positive socio-economic effects this would have.
37
Naturally, there are disparities between the sectors with respect to their productivity levels. Here,
it is worth noting those sectors that see relatively high and low productivity. The utilities sector
(electricity, gas, and water) sees extremely high productivity levels, though mainly due to the little
employment the sector carries. Mining and quarrying carries a similar situation, one of high output
and relatively low employment. Personal services is the first of the larger employment sectors,
with impressive productivity levels, as with finance and manufacturing; the other large
employment sectors. Notably, trade, the second largest employer, sees relatively low productivity,
possibly due to the greater levels of informality in the sector. Further, general government services
sees low productivity levels. Being the largest employer in the economy, it is worth considering
whether government services should be geared towards output, however. Construction, the lowest
productivity sector, sees dismal productivity levels, although this can be attributed to the seasonal
and irregular nature of contractual employment, thus possibly artificially inflating the employment
level of the sector.
Table 6. Sectoral productivity, 2021.
Sector Output (ZAR) Employment
Productivity
(ZAR/emp)
Electricity, gas and water
103 123 757 524
82 000
1 257 607
Mining and quarrying 217 995 833 653 370 000 589 178
Personal services
698 363 579 174
1 258 000
555 138
Finance, real estate and business services
1 054 343 876 281
2 404 000
438 579
Manufacturing 522 884 109 719 1 316 000 397 328
Transport, storage and communication
334 664 582 072
951 000
351 908
Trade, catering and accommodation 522 914 296 189 2 896 000 180 564
Agriculture, forestry and fishing
128 999 033 375
868 000
148 616
General government services
369 419 027 822
3 264 000
113 180
Construction 110 984 428 692 1 133 000 97 956
Source: Statistics South Africa (2022a, 2022b).
Note: Productivity is calculated as sectoral output in constant 2015 ZAR divided by sectoral employment in Q4 2021.
Arranged in descending order.
We may also consider the relationship between employment and output in the South African case.
Whilst the two, generally speaking, have a positive correlation, there are instances, especially in
the periods after crises, that this is not the case. 2009 saw reductions in both employment and
output after the GFC. However, whilst 2010 saw a recovery in output, employment decreased
further. It is only in 2011 was it the case that employment and output grew together. Similarly,
2020 saw a decrease in output and employment from 2019, though whilst 2021 saw a recovery in
output, a further collapse in employment is observed. These cases are similar in their timing and
direction of employment and output, though with different severities and nature of the crises. It is
unfortunate that there is no recent evidence of the response of this relationship in light of “good”
news, though movements between 2003 and 2008 seem to show promise in this respect. A visual
representation of this relationship, in log form, is provided below.
38
Figure 25. Log of employment vs log of output, 2000 – 2021.
Source: Statistics South Africa (2022a, 2022b).
Whatever the chosen route will be for those heading up industrial policy, optimising this
relationship between output and employment is crucial. There is cause for concern of jobless
growth in the South African economy, a potential reality it cannot sustain given the prevailing
labour market and socioeconomic conditions, and indeed the goals of the NDP 2030. A careful
further analysis of this issue is desperately wanting, though beyond the scope of this paper.
Regardless, it is in this context that South Africa finds itself: An ambitious emission reduction
target straddled by political hamstringing and an ailing monopoly; a legitimate and growing
demand for adequate social protection measures in a tight budgetary environment; and
employment targets that have long been abandoned in a labour market that is unable to provide for
its constituents. These issues are severe and crucial to interrogate in a manner that does not
compromise one for another; there is little room for large mistakes in this regard. Addressing these
areas sustainably requires thinking about how they relate to each other, their embeddedness, and
what the driving forces and opportunities behind them are. It is this task that this paper sets to map
out, with a preliminary analysis of these macroeconomic dynamics.
39
4. Methodology
Linking the dynamics of our areas of interest as outlined by an approach of strong sustainability,
the methodology employed makes use of various macroeconomic equations and indicators in
attempt to link factors of growth, emissions, employment, social protection, public debt, and the
balance of payments.21
Ecological perspective:
The Kaya Identity22 (1) is employed as a starting point, being:
=
.
.
.
Whereby is carbon dioxide emissions,23 E is energy (kJ), Y is GDP in constant prices, and
POP represents population. We take the derivative to get this in terms of growth rates (2):
= + + +
In which g is the growth rate, with subscripts CO2 being CO2 equivalent emissions, COint for CO2
intensity (CO2 emissions per kJ produced), Exerg for kJ used to produce one unit of GDP, Ycap is
GDP per capita, and Pop is population.
Here, and are given. Hence, for a given path of CO2 emissions, such as those
supplied by the South African state in their updated Nationally Determined Contribution (Republic
of South Africa, 2021b), we have a constraint of , given the dynamics of the remaining
variables.
Employment perspective:
Simply, if we assume that the growth rate of employment is a function of production (GDP) and
labour productivity, we are left with (3):
=
Where g represents growth rate, and subscripts e, Ycap, and prod denote employment, GDP per
capita, and productivity respectively. Assuming that labour productivity is technologically given,
it is left to the growth of output per capita to determine employment growth. In the results section,
this relationship is augmented slightly in line with historical observations of the relationship
between GDP per capita growth and employment growth.
21 Based on preliminary work by Godin (2021). Strong sustainability in the long-run: squaring the circle?
22 For more information, see https://ourworldindata.org/emissions-drivers
23 Normally, it is GHG emissions that is used. Considering South Africa’s high CO2/GHG ratio and that NDCs are in
terms of CO2, we use it here. This can be easily reverted for studies in other contexts.
40
Social protection perspective:
Cost estimates of improved social protection coverage follows the methodology of The Institute
for Economic Justice (2021),24 updated to 2022 figures.
This follows a focus on coverage and adequacy of social assistance via cash transfers, focusing on
the Child Support Grant (CSG) and the introduction of a Universal Basic Income Guarantee
(UBIG).
Cost estimates are made in increasing the adequacy (level) of the CSG, whilst introducing a UBIG
in the South African context. The CSG is increased to the FPL in 2022, whilst adjusted upward by
forecasted inflation and demographic changes to the age cohort.
The UBIG considers introductions for those between the ages of 18 and 59, making use of various
poverty lines, the Food Poverty Line (FPL), the lower-bound poverty line (LBPL) and upper-
bound poverty line (UBPL), for those not formally employed. Based on experience of uptake in
other major cash transfer programmes, here, we make use of an uptake best estimate of 70% of
those not formally employed between the ages of 18 and 59, increasing at the rate of forecasted
inflation. The 70% is slightly higher than the uptake of a similar though temporary measure, though
accounts for a greater ease of access. These additional costs are then aggregated as a total
additional social protection expenditure relative to GDP.
As such, annual UBIG cost estimates are calculated as (4):
=
( )12
Where is annual cost of the UBIG,
is working age population (18 to 59), Emp is total
employment, formal is the proportion of those employed in the formal sector, is the UBIG
amount, and 12 representing 12 months.
Further, total estimated CSG expenditure is represented as (5):
=
12
Where is annual cost of the CSG,
is child population (0 to 17), is the coverage
rate of those receiving the CSG, is the CSG amount, and 12 representing 12 months.
Balance of payments perspective:
Represented simply, though with space to provide greater specificity (Bhering et al., 2019), the
balance of payments constraint can be represented by (6) and (7) below:
= ++
= ++
24 A methodology developed and paper written by the author.
41
Here, the growth rates of local and global import propensity is and , import price inflation
is given by , export price inflation is , and and are GDP growth rates in the South
African and global economy respectively. With this, we are able to compute domestic GDP per
capita growth restrained by the BoP constraint (8):
= +( )+( )
Public debt perspective:
Following a simplified version of Behring (2021), public debt constraints are dependent on the
debt-to-GDP ratio. The derivative of which (growth rate) is given below (9):
=
Wherein subscript d refers to the debt-to-GDP ratio, subscript D refers to public debt, and subscript
Y refers to output growth. Further, is domestic inflation.
Taking the rate of growth of debt being equal to
, where the deficit is
government expenditure (G) plus debt interest payments (i), less tax receipts (T), we are left with
(10) and (11):
= +.
=
.1
+
As such, a given deficit as a ratio of GDP provides a growth rate of GDP per capita with a stable
debt-to-GDP ratio of (12):
=
.1
+
Where G is inflated by the costs of transition paths and additional social protection spending, with
a fiscal multiplier of 1.4.
42
5. Results and Discussion
Having contextualised the main trends and position of the emission and energy, social protection,
and labour market environments in South Africa, a diagnostic and deconstruction of the dynamics
between these areas is possible. What follows is the implementation, deconstruction, and
discussion of these dynamics, with the aim of gaining a clearer understanding of where the main
frictions and opportunities exist in addressing the advancements of these areas in a holistic manner.
5.1. Emission reduction and output perspective
The growth paths of CO2 emissions and GDP per capita are inextricably linked as we construct
equation (2). Recall that this relationship is such that:
= + + +
Or, that with given or fixed proportions for gExerg, and gPop, we are left with gCOint dependant
on an emissions growth, gCO2, trajectory inline with that of given targets. We are then able to
deduce the growth of GDP per capita that would result in the meeting of these emission targets.
Taking equation (2), as well as South Africa’s NDC targets for 2025 and 2030, we make the
following assumptions:
1. linearly follows the upper and lower-bound paths of the NDC targets, with separate
rates from 2020 until 2025, and 2025 to 2030.
2. is set to zero.
3. is set to zero.
As a result, we are left with the following results, summarised in table 7 below.
Table 7. GDP per capita growth rates, following NDC targets.
Period
High
Target
Low
Target
gCO2 High
gCO2 Low
gYcap High
gYcap Low
2022-2025
510
398
2,45%
-2,51%
1,28%
-3,60%
2026-2030
420
350
-3,81%
-2,54%
-4,70%
-3,52%
Source: Authors construction.
Note: High and Low Targets represent the upper and lower-bound ranges of the updated NDCs, in millions of tonnes
of CO2. gYcap (growth in GDP per capita) and gCO2 (growth in CO2 emissions) are denoted in percent, with High
and Low suffixes referring to the upper and lower bound NDC target paths. gYcap and gCO2 are average annual
growth rates for the corresponding period.
As would be expected, CO2 emission growth rates between the upper and lower-bound targets of
the NDC targets are vastly different. Due to South Africa currently having CO2 emissions below
that of the upper-bound range of the 2025 target, there is space to increase emissions by 2.45% per
year from 2022 to 2025, resulting in positive GDP per capita growth. However, this includes the
3.63% increase in GDP per capita already realised in 2021. Excluding this from the average rate
43
of GDP per capita growth, we see that prospects for GDP per capita growth, in adherence to this
emission reduction path, is slightly lower but still positive at 1.28% per year. Thereafter, there is
a harsh reduction in the GDP per capita growth constraint implied by the sharper slope of decline
of emissions to meet the target of 420Mt of CO2 by 2030, resulting in an average annual constraitns
of GDP per capita growth rate of -4.7% for the latter period. In contrast, the path of GDP per capita
growth for the lower-bound NDC target is smoother, though carries a negative constraint
throughout the 2021-2030 period. Omitting 2021, there is an average annual reduction in GDP per
capita growth of 3.6% from 2022 to 2025, thereafter -3.54% from 2026 to 2030. This is due to the
more ambitious target of the lower-bound NDC target of 398Mt CO2 by 2025 and 350Mt CO2 by
2030. The cases show the varying GDP per capita constraints associated with the different growth
paths necessary to meet the upper and lower-bound NDC targets. Whilst the upper-limit sees
modest GDP per capita growth from 2022 to 2025, there is a sharp and sustained reduction
thereafter until 2026. In following the lower-bound target emission growth path, the GDP per
capita constraint is negative throughout, though relaxes marginally in the latter period.
However, this is likely not the future of actual GDP per capita growth in South Africa. The 2022
Budget Review (National Treasury, 2022b) forecasts real GDP growth of 2.1% in 2022, with 1.6%
and 1.7% in 2023 and 2024 respectively. Accounting for population growth, this corresponds to
figures of GDP per capita growth of 0.91%, 0.45%, and 0.59% for over the medium term (2022-
2024). Considering equation (2), this leaves the viability of the NDC targets in jeopardy without
changes in those variables with some dynamic capacity. Specifically, the growth rates of CO2
intensity (CO2 per PJ energy produced) and exergy (energy per unit of GDP) are of interest.
CO2 intensity is constructed as the ratio between CO2 output and energy produced. There are thus
four ways of reducing this to allow for more “space” for GDP per capita growth:
1. A reduction in CO2 emissions.
2. An increase in energy produced.
3. A greater increase in energy production relative to CO2 emission.
4. A greater decrease in CO2 emissions relative to energy production.
There is room here to increase energy production with existing infrastructure, with reports from
Eskom that there is a gap between desired and actual generation efficiency (Department of Energy,
2019; Sibembe, 2019). This is a potential area of emission reduction relative to energy generation.
Naturally, the obvious solution to the question of decreasing CO2 intensity is the use of renewable
energy generation sources. This is due to their power to supplement existing, high CO2-emitting
coal-fired sources. As such, a greater share of renewable sources in the energy mix can decrease
CO2 intensity. The extent to which this is possible in the short term, considering infrastructure
constraints and the need to service already struggling peak loads, is uncertain.
Further, a deconstruction of exergy is useful in this analysis. Exergy, as the energy used per unit
of GDP, will decrease with an increase in GDP (and GDP per capita) given a fixed amount of
energy consumption. As such, it may be such that annual GDP per capita growth higher than the
figures in table 7 may offset associated increases in the growth rate of CO2 emissions. However,
the composition of what constitutes this energy use is not addressed. In a case whereby increases
44
in energy generation occur, the carbon profile of this additional energy generation matters. If, in a
scenario whereby renewable sources are deployed in addition to existing sources, an increase in
energy production would result. However, it does not necessarily follow that the resulting increase
in exergy would increase the growth rate of CO2 emissions. It is here that equation (2) requires
more nuance.
What is obvious for the dual targets of emission reduction as well as output growth is that their
dynamic is linked. Historically, we have seen a strong correlation between increases in emissions
and output. Due to South Africa’s carbon intensive economic structure, this is to be expected. Now,
however, there is a trade-off to be made between emission reduction and “development” in lieu of
an unchanging energy generation mix and technological standing. For contemporary South African
policymakers, there is no deliberation as to which side of this dichotomy is preferred. The point of
equation (2) is to deconstruct this dichotomy, however, and show that there is room to manoeuvre.
This requires a deep and concerted effort, however, with strong institutional capacity, something
that both the ANC and Eskom have been found wanting in for at least the last two decades.
Equation (2) thus highlights the imperative to change the energy mix of the South African
economy, away from coal, in order to be able to accommodate both higher output growth and
emission reductions in line with the NDC targets.
We may consider, then, how table 7 changes with changes to both CO2 intensity and exergy levels.
More realistically, it is likely that the introduction of greater renewable energy generation capacity
will be in addition to existing fossil-fuel capacity. As such, we would expect total energy to
increase. Here, the effect would be a greater decrease in CO2 intensity. Further, we are able to
envision a scenario where South Africa improves its exergy levels. As such, we provide an
alternative view of the constraint of GDP per capita growth with the assumptions that:
1. Energy generated is increased by 5% per year.
2. Exergy is linearly improved to 0.55PJ per ZAR billion GDP by 2030.
Table 8. Adjusted GDP per capita growth rates following NDC targets.
Period
High Target
Low Target
gCO2 High
gCO2 Low
gYcap High
gYcap Low
2022-2025
510
398
2,45%
-2,51%
5.66%
5.42%
2026-2030
420
350
-3,81%
-2,54%
5.69%
5.75%
Source: Authors construction.
Note: High and Low Targets represent the upper and lower-bound ranges of the updated NDCs, in millions of tonnes
of CO2. gYcap (growth in GDP per capita) and gCO2 (growth in CO2 emissions) are denoted in percent, with High
and Low suffixes referring to the upper and lower bound NDC target paths. gYcap and gCO2 are average annual
growth rates for the corresponding period.
Here we can see that with the changes in the growth paths in CO2 intensity and exergy, GDP per
capita is given far more room on the upside whilst still, theoretically, being able to meet the
emission targets of the NDCs. To follow the upper-bound NDC target of 510Mt CO2 by 2025,
GDP per capita’s constraint would be an average of 5.66% per year. For the following 5-year
period, a restraint of 5.69% exists. Following the lower path, constraints of 5.42% per year and
5.75% per year respectively exist. This would be a much better situation for South Africa, as it
45
gives room for improved developmental outcomes associated with output growth. Even so, GDP
per capita growth of these magnitudes is not likely to materialise given historical growth rates.
What this means is that with a likely reality of GDP per capita growth beneath five percent, if
South Africa is able to meet the assumptions of energy generation growth of 5% (with renewables)
and improve the efficiency of its energy use, a growth rate of emission reductions beyond that of
the ambitious lower-bound range of the NDC targets is possible. The important change between
table 7 and the previous table 8 is that of the more aggressive (negative) growth paths of CO2
intensity and exergy.25 Given the equation, greater negative growth in these measures allows for
greater room for the GDP per capita constraint to fill that gap (given population growth). Here,
highlighting the importance of improved energy generation capacity, its structure (mix), as well as
improving energy efficiency in producing output (exergy), is essential. These are the areas in which
policymakers should be focussing in order to break the dichotomy of emission reduction and
economic growth.
5.2. Employment perspective
Recall that gains in employment can be defined simply with the relationship set in equation (3):
=
Equation (3) is set such that growth in employment is a function of growth in GDP per capita and
growth in productivity. With results from equation (2) for growth in GDP per capita associated
with those necessary for South Africa to meet its 2025 and 2030 NDC targets, we are able to
estimate the associated growth path of employment. Recall that these results from equation (2) are
disaggregated into two periods, 2025 and 2030, and two trajectories towards the upper-bound and
lower-bound extremes of the NDC targets. Assuming that there is no growth in productivity, whilst
setting it to pre-pandemic levels, we are left with the following results for equation (3):
Table 9. Employment growth associated with table 7.
Period
High Target
Low Target
gYcap High
gYcap Low
gEmp High
gEmp Low
2022-2025
510
398
1,28
-3,60
1,28
-3,60
2026-2030
420
350
-4,70
-3,52
-4,70
-3,52
Source: Authors construction.
Given constant productivity levels (a growth rate of productivity of zero), we find no difference
between the annual average growth rate in GDP per capita associated with the emission growth
rate of meeting the upper and lower-bound NDC targets. The same can be seen for the adjusted
scenario of table 8, with the positive, more relaxed GDP per capita growth constraint enabled by
improving CO2 intensity and exergy.
25 Plots of both scenarios are available in the appendix.
46
Table 10. Employment growth associated with table 8.
Period
High Target
Low Target
gYcap High
gYcap Low
gEmp High
gEmp Low
2022-2025
510
398
5,66
5,42
5,66
5,42
2026-2030
420
350
5,69
5,75
5,69
5,75
Source: Authors construction.
In the first, simple, table 9, we see negative employment growth besides marginal growth in the
upper-bound target in the period 2022 - 2025. In the context of high unemployment, poverty, and
inequality embedded in the South African economy, this is an undesirable future even if emission
reduction targets are to be met in this scenario. The tensions revealed in this framework between
the goals of emission reduction and employment growth are stark. Given the current composition
of the economy, it will not be possible to meet both targets, perhaps forcing policymakers into a
dilemma. It is here that the importance of a holistic view is revealed. Again, it is more likely for
emission reduction targets to be sacrificed than to worsen the already deep crisis of unemployment.
We observe the importance, again, of the need to change the energy generation and emission
structures of South Africa in a well-considered and timeous manner. Further, it may well be that it
is only possible to solve the multiple and strongly linked challenges of emission reduction,
unemployment, poverty, inequality, and social protection coverage and adequacy that any of these
targets become desirable.
A further deconstruction of equation (3) proves useful to our understanding of the GDP per capita
vs employment growth relationship. Firstly, an analysis of this relationship is useful.
Figure 26. GDP per capita growth vs employment growth, 2001 – 2021.
Source: Authors construction, Statistics South Africa (2022b, 2022a).
Note: Blue line is linear regression, green line is loess.
As illustrated in figure 26, the historical relationship between GDP per capita growth and
employment growth is non-linear. With large decreases in GDP per capita growth, such as in 2009
47
and 2020, there are also large decreases in employment growth. Of note is the immediate response
to the year of intense crises, however, with 2010 and 2021 seeing recoveries in GDP per capita
growth without recoveries in employment growth. From experience of the GFC, we see a recovery
in both measures two years after the initial shock (2011), which may lead us to believe that an
employment recovery is likely for 2022. Further, and more salient to the results of GDP per capita
constraints associated with the NDC target growth paths illustrated earlier, we can augment the
relationship between GDP per capita growth and employment growth by way of historic
relationships. For the baseline, no-technological gain scenario, we see that employment growth in
2012 and 2013 of between 2.5% and 3% for GDP per capita growth of 0.8, but lower employment
growth in 2011 with relatively higher GDP per capita growth. As such, for the period 2022 to 2025
with GDP per capita growth of 1.28%, we augment the equation to correspond to employment
growth of 2%. In the following period with negative growth of 4.7%, we assign an optimistic
employment response of 3%. This is due to the knowledge of the shock, whereas the negative
shocks on 2009 and 2020 were unexpected. If contingencies can be made in light of this, we have
reason to believe stronger employment protection measures would be employed, dampening the
negative employment growth shock. Similarly, for the lower-bound NDC target growth paths, we
assign employment growth values of -2% for the corresponding -3.6% and -3.5% GDP per capita
growth figures. Further, for the scenario with technological gains being energy generation growth
and improving exergy efficiency, we see strong employment responses in 2005 and 2006, the
height of the employment gains made during the commodity supercycle period. As such, with
strong GDP per capita growth, we take the upper-limit of these observations and augment the
employment growth response to 6% for both periods and both NDC target growth paths.
With these augmentations, we are able to see what more realistic employment growth might look
like in light of the potential GDP per capita growth scenarios. We are also able to calculate what
the resulting narrow unemployment rate would be given labour force size projections.
Table 11. Augmented employment vs GDP per capita growth scenarios.
Period
gYcap
High
gYcap
Low
gEmp
High
gEmp
Low
Emp
High
Emp
Low
Unemp
High
Unemp
Low
2022-2025
1,28% -3,60% 2,0% -2,0% 15,90m 13,55m 34,0% 43,8%
2026-2030
-4,70% -3,52% -3,0% -2,0% 13,66m 12,25m 48,0% 53,4%
Source: Authors construction.
Note: g_Emp denotes the growth of total employment, with High and Low referring to the upper and lower-bound
NDC targets. Emp total denotes the total level of employment measured in millions at the end of the relevant period.
Unemp refers to the narrow unemployment rate at the end of the relevant period.
We see that with the non-linear relationship between employment growth and GDP per capita
growth, the scenario of negative GDP per capita growth is still negative, though not as severe.
Following the upper-bound NDC target growth path, the (un)employment growth constraint is 2%
per year until 2025, leading to an unemployment rate of 34% if met. Thereafter, this worsens to
48% by 2030. Following the ambitious path, we see employment growth constraints of -2% for
both periods, with corresponding levels of unemployment of 43.8% and 53.4% in 2025 and 2030
48
respectively. These constraints constitute negative absolute employment growth from Q4 2021.
Crucially, the NDP target unemployment rate of 6% by 2030 will not be reached with this level of
employment growth. Here, the constraint on unemployment growth is the constraint of GDP per
capita growth associated with an unchanged energy generation and use structure used to achieve
the necessary negative growth rate in emissions.
With a change in such structure, as outlined previously, both GDP per capita and employment
growth constraints are very different, as highlighted in table 12 below.
Table 12. Augmented employment vs adjusted GDP per capita growth scenarios.
Period
gYcap High
gYcap Low
gEmp High
gEmp Low
Emp total
Unemp
2022-2025
5,66%
5,42%
6,0%
6,0%
18,55m
23,0%
2026-2030
5,69% 5,75% 6,0% 6,0% 24,82m 5,5%
Source: Authors construction.
Note: g_Emp denotes the growth of total employment, with High and Low referring to the upper and lower-bound
NDC targets. Emp total denotes the total level of employment measured in millions at the end of the relevant period.
Unemp refers to the narrow unemployment rate at the end of the relevant period.
In a structure of improving energy generation and energy-use efficiency, a higher GDP per capita
growth rate constraint is associated with meeting the NDC emission targets. Subsequently, we
observe employment growth rates similar to those of the height of employment gains in South
Africa. What results is a sharp reversion of recent labour market trends, whereby the more relaxed
GDP per capita constraints enable total employment to increase to 18.55 million by 2025, and
24.82 million by 2030 if met. With estimated labour force sizes in 2030 to be 26.28 million
(National Planning Commission, 2020), this results in unemployment rates of 23% by 2025 and
5.5% by 2030. This is a massive course correction for South Africa in terms of its labour market
dynamics, and coincidentally almost exactly the 6% target unemployment rate outlined in the NDP
2030.
The elegance of equation (3) is that it allows for a reversal in the direction of analysis, such that a
target employment growth rate can be used to determine the necessary GDP per capita growth,
which can be substituted into equation (2) to garner an idea of what the resulting emission growth
path constraint would likely be (given the dynamics of the other variables). However, without a
more detailed analysis of the relationship between employment growth and GDP per capita growth
on a sectoral level, these estimates leave room for improvement. The aim here however is not to
provide very accurate estimates, but to interrogate the dynamics between emissions, growth, and
employment. Evidently, the dynamics are linked. Further, constraints of emission growth, GDP
per capita growth, and employment growth can be changed, such that one of the measures is given
“priority”, and the remaining measures seen as the technical constraints towards achieving that
target.
Further, we may wish to analyse the effect that productivity has on potential employment. As
expected, a decrease in productivity leads to greater scope for employment growth given a GDP
per capita constraint. In simple terms, that greater increases in employment than in output lead to
49
a greater amount of people able to be absorbed into the labour market. With labour intensive
employment growth, this is possible. We see this relationship generally present in the South
African economy too, with historical evidence in figure 26 showing that employment growth tends
to outstrip GDP per capita growth. In this case, we may find an even more relaxed employment
growth constraint with an increase in GDP per capita. However, it is important to remember that
the associated employment growth constraint is the maximum constraint, and not necessarily the
given future of employment given some level of GDP per capita growth constraint. In any setting,
sustained employment growth of 6% would be very impressive. In South Africa’s case of relatively
stagnant employment growth and a steadily increasing unemployment rate since 2015, this is
unlikely to materialise. We are then able to conclude that the constraint of employment growth
will likely not be salient to the feasibility of achieving the emission growth path set out by the
NDCs under this scenario.
5.3. Social protection spending and macroeconomic linkages
Section 3.2. included a basic costing of improving and expanding social protection measures for
two groups. Children aged 0 to 18 years of age are to receive an increased level of monthly support,
and the working-age population are to receive support sufficient to survive. We reproduce the total
additional spending below:
Table 13. Additional spending for increased social protection coverage (ZAR bn), 2022-2025.
2022
2023
2024
2025
CSG @ FPL
22,81
24,04
25,19
26,38
UBIG @ FPL, 70% NFE
128,65 134,82 140,75 147,09
Total additional
151,46 158,86 165,94 173,47
Source: Authors construction.
In order to contextualise the impact of this additional spending within the current and projected
level of social protection spending,26 we turn to figure 27 below:
26 Projected spending on social protection by National Treasury uses 2024 as the most distant estimate, hence the
inconsistency in the dates between table 13 and figure 27.
50
Figure 27. Total and Social Protection Expenditure, expanded, 2005/06 – 2024/25.
Source: Authors construction. National Treasury (2022).
Note: SP refers to social protection, which includes measures in addition to social grant expenditure. SP + refers to
the additional expenditure as per proposals. Total + refers to total spending inclusive of additional proposed
expenditure. Ratio + refers to the ratio with increased spending of proposals added to SP and Total. Left axis is
spending, right axis is ratio in %.
As is evident, the proposed measures work to reverse the downward trend in social protection
spending as a portion of total spending that is planned for 2022 to 2024. The introduction of these
measures move the projected proportion of social protection spending to total consolidated
expenditure from 17.8% in 2022 and 15% in 2024 to 23.3% and 21% in 2024, about 6% higher
with the new introductions as a proportion of total spending, translating to a total spend on social
protection of R521 billion in 2022 and R494 billion in 2024 (vs current forecasts of R370 billion
and R328 billion respectively). As mentioned, the current forecast of the state reveals an
anticipation of, in a gratuitous reading, a decrease in demand for social protection, i.e., that people
will be better off and in less need of state support. However, this is difficult to reconcile with the
recent impacts on the labour market due to COVID-19. With job-losses in the millions, and no
structural changes with regard to increased capacity for labour absorption beyond pre-pandemic
levels, demand for state support will likely be higher than pre-pandemic levels. How this translates
to a reduced social protection spending requirement is unclear.
Further, we note that social protection demand is primarily linked to poverty. With the labour
market bearing structural dysfunctionalities, it is imperative to reform this in order to improve
labour absorption, inequality, and fiscal pressure due to increasing social spending requirements.
It then follows that social protection spending hinges partially on the labour market’s prospects
and ability to bring a greater part of the working age population into its benefits. In light of the
measures proposed, a more careful analysis of qualification criteria for receipt of social protection
measures has not been conducted. For simplicity’s sake, we are able to deduct the following: that
social spending requirements are reduced by greater labour market inclusion. For the Child
Support Grant, larger average household incomes would lessen the need for state support,
0
5
10
15
20
25
-
500,0
1 000,0
1 500,0
2 000,0
2 500,0
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
SP/Total (%)
Spending (ZAR bn)
SP Spending SP Spending + Total Spending
Total Spending + Ratio + Ratio
51
decreasing social spending requirements. As it stands, with some 65% of children in South African
benefitting from this measure, this is incredibly important. For the Universal Basic Income
Guarantee, as defined here as 70% of those not formally employed being eligible for receipt,
greater labour market participation decreases the number of people who cannot support
themselves, and thus decreasing the spending requirement necessary to fulfil this Constitutional
mandate.
Given the composition of the additional expenditure required by these introductions, we see that
it is the UBIG that comprises the bulk of the cost. However, this cost hinges on the labour market,
under which context provides some of the demand for social protection cover. We may then
analyse changes in expenditure required given levels of employment. We use the two scenarios of
GDP per capita and employment growth constraints from sections 5.1. and 5.2. in order to estimate
possible changes in expenditure requirements of the UBIG, with the annual results as follows.
Table 14. Spending requirements of a UBIG, 2022 – 2030.
Year
formal_pop
(%)
formal_gr
(%)
NFE_pop
high (m)
NFE_pop
low (m)
NFE_gr
(m)
Amount
(R)
Total_pop
high (R b)
Total_pop
low (R b)
Total_gr
(R bn)
2022
71,9
72,8
17,0
17,3
16,6
624
127,4
129,6
124,5
2023
71,9
73,7
17,2
17,8
16,3
654
134,8
139,4
128,3
2024
71,9
74,6
17,3
18,2
16,0
683
142,0
149,3
131,4
2025
71,9
75,5
17,5
18,7
15,7
713
149,9
160,0
134,5
2026
71,9
76,4
18,1
19,1
15,3
746
161,6
171,2
136,9
2027
71,9
77,3
18,6
19,6
14,9
779
174,1
183,2
139,0
2028
71,9
78,2
19,2
20,1
14,4
814
187,4
196,1
140,6
2029
71,9
79,1
19,7
20,5
13,9
851
201,5
209,6
141,5
2030
71,9
80,0
20,3
21,0
13,3
889
216,4
224,0
141,4
Source: Authors construction.
Note: formal denotes the percentage of total employment in the formal sector, NFE denotes 70% of those not formally
employed, Amount denotes the amount of the monthly grant, Total denotes total annual cost. Suffixes _pop and _gr
refer to the baseline and adjusted GDP per capita growth and employment growth constraint scenarios from section
5.1. and 5.2., reflective of constant and improved CO2 intensity and exergy respectively. High and low refer to the
upper and lower-bound NDC targets respectively.
Table 14 highlights the differences in expenditure requirements that follow changes in
employment growth under our two GDP per capita growth constraint scenarios. We further
augment the percentage of those employed in the formal sector under the higher growth scenario
to linearly increase to 80% of total employment by 2030. With employment growth as much as
6% per year, this is likely to, in part, come in the form of state employment programs and
expansions of already large employment share sectors. Regardless of the form this will take, we
aim to show that an increase in the share of those employed in the formal sector, under the given
qualification criteria of the UBIG proposed, reduces the number of those in need of state support.
This has a dual benefit of bringing more people into the formal sector with greater legal protection,
higher wages, and less precarity, and also reducing the state expenditure requirement of a UBIG.
We inflate the amount of the monthly benefit by forecasted inflation until FY2024/25, and 4.5%
thereafter.
We are left with diverging expenditure requirements necessary for the UBIG. In the negative-
growth scenario, the requirement increases from R127 – R130 billion in 2022 to R216 - 224 billion
52
in 2030, a 70% increase. In contrast, the high growth constraint scenario sees increases of just
R16.9 billion over the 2022-2030 period, a 14% increase. Clearly, the most important factors
deciding the demand (cost) of the social protection coverage for the working age population are
that of those employed and the proportion of those employed in the formal sector. Increases in
both of these measures reduce the demand and hence expenditure requirement of this portion of
closing the social security net.
What we do not anticipate further is the result of changes in employment or the introduction of a
UBIG on the demand for the proposed increase in the Child Support Grant. For simplicity’s sake,
we keep the forecasted increases in expenditure unchanged in both scenarios of negative and
positive GDP per capita growth constraints. However, it is almost certain that with an increase in
the number of those employed to the extent of the high-growth scenario will reduce the demand
(number of those qualified by the current means testing method) of the CSG and hence its
expenditure requirements. Finally, total forecasted increases in social protection spending is
provided below.
Table 15. Total increases in social protection funding, 2022 – 2030.
Year
+CSG (R bn)
Total_pop high (R bn)
Total_pop low (R bn)
Total_gr (R bn)
2022
22,8
150,3
152,5
147,3
2023
24,0
158,8
163,4
152,3
2024
25,1
167,1
174,4
156,6
2025
26,0
175,9
186,1
160,5
2026
27,4
188,9
198,6
164,3
2027
28,6
202,7
211,8
167,7
2028
29,9
217,3
225,9
170,5
2029
31,1
232,7
240,8
172,6
2030
32,5
248,9
256,4
173,9
Source: Authors construction.
Note: +CSG denotes additional spending on the CSG to increase it to the FPL. Total_pop denotes total additional
spending on social protection in the negative GDP per capita constraint scenario, Total_gr denotes total additional
spending in the increased/high GDP per capita growth constraint scenario. High and low refer to the upper and lower-
bound NDC targets respectively.
5.4. Public debt constraints and dynamics
In light of muted output growth over the last decade, South Africa has seen an increasing debt-to-
GDP ratio. This is often cited as a reason for reduced funding for developmental outcomes by
National Treasury. Financing requirements for both increased social protection coverage and to
fund the infrastructure development necessary to lower emissions at the NDC-implied rates imply
further pressure. With the small tax-base that comes with a dysfunctional labour market, the state
has to carefully consider debt financing options over and above its existing commitments in lieu
of broader structural changes to the tax system and productive and (re)distributive systems.
Financing the shift away from coal dependency has been contentious. Already outlined in Section
3.1., the conditionalities and rates that come with multilateral funding has not been clearly laid out
53
for analysis. However, there are some estimates of what level of funding will be necessary to make
these changes. The International Energy Agency (2020) estimates a cumulative figure of some
$147 billion of investment in renewables and their associated networks from 2019 to 2040, crudely
reflecting an annual investment of R111 billion per year from 2022 to 2030. Additionally, the
updated September 2021 NDC document (2021b) highlights the need for some R105 billion a
year.27 However accurate, we can speculate that the annual cost should technically be in the region
of R105 billion a year in order to meet the lower-bound target of the NDCs.
How these measures above will affect public finances and debt constraints is a crucial
consideration. Recalling equation (12):
=
.1
+
We see that for a given GDP per capita growth path, such as the ones provided in different
scenarios prior, we are able to determine a debt-to-GDP ratio growth rate constraint. For this, we
run 6 scenarios, two for each non-technological gain scenario and two for the technological gain
scenario, with the following assumptions:
1. Additional expenditure on expanding social protection is funded solely through debt.
2. Additional financing estimates of R105bn per year are required to meet the lower-bound
NDC target, and R32bn per year for the upper-bound target, and are funded through debt.
For the technological gains scenario, we take the midpoint of these annual costs.
3. Revenue receipt as a proportion of GDP increases linearly to 26% by 2030.
4. Non-additional expenditure increases inline with the average increase from 2020 to 2024.
5. Interest payments on debt are made at 6.5% from 2026 onwards.
It is here that we calculate the two scenarios for each iteration of no and high-technological gains
and their associated GDP per capita constraints. The first scenario has to do with the multiplier
effect of this additional expenditure. There has been fervent recent debate on the literature of
multipliers in South Africa for at least the last three years, with the methods employed in its
estimates of further debate. Some have attempted to show that expenditure multipliers are zero,
for example by studies from the National Treasury and the Reserve Bank (Studies in Poverty and
Inequality Institute, 2021). It is for this reason that we provide two scenarios with different
multipliers. From what is the most likely fiscal multiplier,28 we consult Schroder & Storm’s (2020)
work, which estimates fiscal multipliers for direct, indirect, and induced consumption effects of
1.5 (and employment of 6.1, for interest). To temper this towards a more conservative effect, we
provide the effects of debt-to-GDP ratio growth for each scenario, disaggregated by fiscal
multiplier effects for additional expenditure of both 0.8 and 1.4.
However different the multipliers are from each other, however, we find no significant difference
in terms of the effects that it has on the equation
, as any effect is drowned by the far stronger
27 Midpoint estimate between R860 and R920 billion in total.
28 Due to the relative rigour, most recent datasets used, and more realistic assumptions.
54
changes in output associated with GDP per capita constraints. As such, the results of only the 1.4
times fiscal multiplier are presented below.
Table 16. Debt-to-GDP ratio constraints.
Year
g_Ycap High
g_Ycap Low
g_Ycap Gr
g_d High
g_d Low
g_d Gr
2022
1,32
-3,64
5,55
16,16
24,16
12,07
2023
1,30
-3,65
5,63
16,32
25,50
10,70
2024
1,28
-3,68
5,70
15,08
25,27
8,03
2025
1,25
-3,72
5,77
15,83
26,94
7,33
2026
-4,57
-3,46
5,56
21,82
26,13
4,74
2027
-4,67
-3,48
5,62
22,94
26,49
2,94
2028
-4,78
-3,52
5,69
24,00
26,84
1,19
2029
-4,90
-3,55
5,76
25,00
27,20
-0,55
2030
-5,04
-3,60
5,82
25,96
27,55
-2,30
Source: Authors construction.
Note: g_Ycap denotes growth in GDP per capita. High and Low denote the scenario of no technological gains of the
upper and lower-bound NDC targets respectively. g_d denotes growth in the debt-to-GDP ratio. Gr denotes the
technological gains scenario.
As evident, the scenarios with negative GDP per capita constraints (blue highlight) are untenable
in terms of growth in their associated debt-to-GDP ratios. However, this is to be expected in a
context of declining tax revenues and output, but significant increases in expenditure in order to
expand social protection coverage and enable a transition away from coal-dependence. The results
here start with debt-to-GDP ratio growth constraints of at least 16% and 24% a year, converging
by 2030 to some 26% to 28% increase. In contrast, the scenario with technological gains in CO2
intensity and exergy see manageable debt-to-GDP per capita growth constraints, such that they are
‘desirable’ given their assumed socioeconomic effects. With the above figures, we should interpret
them with greater nuance. Here, it is that if operating at the maximum of the GDP per capita
constraint, there is an associated growth rate of the debt-to-GDP given the assumptions and
scenarios. In light of National Treasury’s stance against increasing an “already high” public debt-
to-GDP level of some 65%, the associated increases in double digits are unlikely to materialise in
in order to achieve the goals set out by the NDC or for the improvements to social protection
coverage and adequacy.
Differences emerge here as to where the defining factors of the growth rate of the debt-to-GDP
are. Firstly, there is the assumption that all additional spending over and above the forecasted
existing non-interest expenditure is debt funded. This may not necessarily be the case in terms of
financing the ecological transition, where financing may be highly concessional. To this point,
there is also the interest rate paid on existing and new debt that is salient. A lower rate would
decrease this growth rate. The South African Reserve Bank is able to make adjustments here,
should they choose to, which would decrease this. The majority of South Africa’s debt is also
denominated in local currency, and as such carries limited depreciation risk. If possible, financing
agreements should be made in local currency in order for the SARB to intervene in this regard.
Further, there are no additional taxes included in the equation and calculation. This may prove
inaccurate and surely undesirable, as tax introductions and adjustments such as Social Security
55
Tax, Resource Rent Tax, a luxury value-added tax (VAT), and a wealth tax have been proposed
by civil society groups and have gained traction in policymaking circles (IEJ, 2021).29 The
introductions of additional revenue raising mechanisms would decrease the debt-to-GDP growth
restraint by reducing the fiscal deficit. Over and above this, the introduction of a UBIG would
likely “pay for itself” to some extent due to consumption effects through VAT (Institute for
Economic Justice, 2021). In addition, the effects on inflation of such a shock of negative GDP per
capita growth could well decrease inflation and inflation expectations, loosening the debt-to-GDP
constraint. Above, we keep those expectations of the National Treasury, though this is a
consideration that could meaningfully impact the results obtained.
At this point, we are also able to abstract away from specific results. From the observations in
results sections 5.1., 5.2., 5.4., and now 5.4., we see what are two stylised paths. The first, under
no gains in CO2 intensity and exergy, is that of a severe and unrealistic trade-off between meeting
the upper and lower-bound targets of the NDC emission reductions versus employment targets,
political and fiscal viability of sufficient social protection, and an untenable public debt position.
The other involves gains in energy generation and output efficiency, with the possibility of solid
employment growth, and a public debt position that would not induce overburdensome constraints
towards expanded social protection measures nor the financing required to meet emission
reduction goals. It is obvious what is more desirable. However, we have further shown that there
are various mechanisms by which to accommodate these goals in light of a framework of strong
sustainability. It is here that this paper aims to synthesise and outline the possible futures of South
Africa’s environmental, social, and economic landscapes, without specificity, in the context of an
already dire situation.
5.5. Balance of payments constraints and dynamics
Lastly, there is a consideration of balance of payments constraints to consider. We cannot
meaningfully run simulations of equation (8) due to the magnitude of changes that would come
with the transition towards lower emissions growth rates and associated constraints, though we are
able to conduct a brief analysis of its composition. Recall equation (8) being:
= +( )+( )
Where is growth in GDP per capita, is global output growth, and are price inflation
of exports and imports respectively, and and are growth rates of import propensities of the
global and domestic economy respectively. is growth in local population.
We see that for local constraints to GDP per capita growth hinge on a few important dynamics.
Firstly, positive and increased growth in global output would create a more relaxed GDP per capita
constraint. Further, greater price inflation of exports relative to imports would have the same effect,
as would a larger growth rate of global import propensity relative to the local iteration. Further,
there is the effect that currency plays, as import growth rates are measures in common currency,
29 From personal involvement in trilateral meetings and presentations.
56
usually foreign, so as to balance. As such, the relative strength of the South African Rand (ZAR)
is an important consideration.
On the question of currency movements, ZAR movements are generally not determined by local
action in the short term, but rather global risk sentiment. In the medium and long-term, credit
ratings as well as commodity price trends drive appreciation or depreciation of the rand. With this
in mind, South Africa’s debt position and current account balance are likely to be impacted by the
manner in which a transition away from coal-dependence takes place. On the one hand, negative
GDP per capita growth associated with negative emission growth will put pressure on revenue
receipts and drive fiscal deficits upwards. How the state decides to cover this deficit is important,
as it has a direct impact beyond the obvious on the balance of payments. Should this be difficult
to manage, and in this scenario that is almost certainly the case, it jeopardises the transition (as
well as the closing of the social protection net). An appreciation (relative weakening) of the ZAR
due to a more severe debt position and wider deficit in a context of negative growth makes the
transition less likely still, due to its effect on the cost of capital goods necessary to reduce
emissions. Here lies another important consideration – the impact of intellectual property transfer
(Chandrasekhar and Ghosh, 2022) restrictions on South Africa’s import propensity. Should South
Africa need to import the input materials and end-products of renewable energy infrastructure, a
weaker rand increases this, putting a greater strain on the ‘room’ for upward GDP per capita growth
as per equation (8) due to the relative increase in import price inflation and propensity to import.30
Naturally, the price effects are more likely to outweigh the import propensity effects, as a weaker
rand affects the whole sphere of imports , whereas transition-specific import propensity increases
may be offset by this same import price inflation.
Conversely, there is an opportunity for more relaxed GDP per capita constraints from a balance of
payments perspective. Should IP be easily accessible, as it ought to in an environment of unmoving
environmental boundaries,31 South Africa may be able to manufacture the necessary capital goods
necessary for its transition. In this case, it is uncertain as to the extent of upward movement of
local import propensities. This would be a brighter scenario, with the potential for South Africa to
help manufacture and install low-carbon technology in the Southern African sub-region where
technical capacity is limited, improving the balance of payments perspective.
Where an analysis of the balance of payment constraint would benefit is through means of an
industry level diagnostic. Here, the effects of minerals and mining sectors would see in light of a
shift away from coal would be valuable as it would affect export volumes and prices, as well as
indirect, downstream production and export effects of other industries. Alternatively, we may
hypothesise that with lower local use of coal, exports of the material would increase substantially.
In terms of emission reductions this would be counterproductive in a global sense, emission
reduction effects need to be global and not shifted from one country to the next, but would bolster
30 Given constant or relatively lower export price inflation.
31 Though still unlikely, as seen with the IP associated with other, more recent health crises and their mitigation
strategies.
57
foreign currency earnings deficits that may be required to pay for the necessary manufactured
capital from abroad.
As illustrated, the trajectories of the effects of the transition on the balance of payments are widely
varying. It is for this reason that a more concrete indicative analysis of the constraint of the BOP
on GDP per capita has not been conducted. However, the dynamics above are those which can
already be considered by policymakers in planning and compensating for transition effects on the
wider macroeconomic position of South Africa.
5.6. Political implications and considerations
The futures of a potential transition away from coal in order to meet the target CO2 emission
trajectories as outlined by the September 2021 update of the Nationally Determined Contributions
leaves room for manoeuvre. As shown, there are ways in which policymakers can meaningfully
impact the constraints involved with meeting these targets, though in the most optimistic picture,
this would require an unprecedented level of cooperation, implementation success, and timeliness.
Above, we start the analysis of the various constraints on the basis that emission reductions are the
priority target, such that all other dynamics are considered to accommodate that target being met.
However, contemporary political economy in South Africa indicates that this target will not be the
priority. Employment growth and GDP growth, of the constraints considered, is probably the area
of most concern for the state and its constituents. Fortunately, due to the flexibility of the dynamics
between the constraints, it is possible to assess what level of emission reduction is possible in line
with given targets in these areas.
However, what is important to consider is the nature of these targets and constraints. In some
instances, there is room to negotiate, although difficult and contentious: some compromises. In
others, there is no room for negotiation, and we can view these constraints and targets as “hard”
constraints. There is no negotiating table at which to sit with emission reduction and environmental
degradation, for example. In contrast, there are formal lines of negotiation in place to revise
“acceptable” levels of the debt-to-GDP ratio, or how to extend social protection to those that need
it. What we may speculate on is that the two technically hard constraints of those considered are
those of emission reduction growth and employment growth. Emission reduction is essential for
two reasons, inter alia. First and most obvious is that it the failure to do so endangers the physical
environment within which all else, the social and the economic, operates. Reducing reductions in
the global context will mitigate climate volatility and increased social and economic vulnerability.
In the local context, reducing further direct environmental degradation allows those living in
affected areas to lead healthier lifestyles with benefits in all aspects. Secondly, is that it provides
an opportunity to restructure a failed labour market and productive structure. South Africa has, for
at least the last decade, been in the midst of a tailspin in terms of social and economic indicators,
with little effect of course-correction plans. The incidence of an ecological transition presents an
opportunity to take stock of where these deficiencies are and work towards correcting them with
aims at more equitable and inclusive distribution from productive activities. This is directly linked
to labour market considerations, whereby the already high unemployment rate is cause for major
concern. A partial restructuring of sectoral employment will likely be necessary for the transition
58
to happen. This is because of the material needs of the transition, as well as the effects that it has
on upstream and downstream output (Espagne et al., 2021). As with previous and existing
industrial policy plans, depressed employment levels would benefit from increased labour
intensity; a larger share of labour relative to capital inputs to production. We may view this
necessity also in light of its inverse, the prospect of allowing unemployment to slide further
upward. Whilst this is technically possible, it is highly undesirable, and further worsening of the
labour market environment would indeed test the limits of an already fragile social compact.
However, there remains an important distinction between what is technically possible and what is
politically viable. Here, constraints that come with an unchanging technological structure are
overly burdensome. We can assume that in light of a negative GDP per capita and employment
growth path and high debt-to-GDP ratio increases that emission reduction targets will be
abandoned. In order to improve outcomes, deep and wide consultation of stakeholders is crucial,
including government (both in presence, and within and between departments), National Treasury
and the South African Reserve Bank specifically, business leaders, workers, workers
representatives in affected sectors, local communities, and civil society groups. The National
Economic Development and Labour Council (Nedlac) would be an appropriate avenue of
consultation in this regard, though insufficient in terms of stakeholder representation. It is here that
further frictions are likely to be identified, though also a space where compromises and salient
strategies can be developed with the necessary input and buy-in from those with decision-making
power in their respective spheres. Naturally, this will involve difficult decisions. Especially in the
realm of energy generation methods, hard political stances will have to be made in the context of
the plurality of visions of emission growth. Whether the state is willing and able to walk the talk
on its NDC targets and other commitments will obviously be the deciding factor in whether a
transition can happen. This will, as discussed, require a break from hegemonic economic ideology,
however, though one that has not proven fruitful in light of any of its developmental targets or
otherwise.
6. Limitations and further work
A range of limitations exist in the results discussed in section 5. These are briefly discussed,
followed by considerations for future work in developing the topic and methodology employed.
Firstly, they analysis makes use of a range of assumptions. Whilst forward looking, this is
necessary to a degree as it is impossible to know the future values of some indicators. However,
in some instances these assumptions are made somewhat arbitrarily in order to illustrate different
growth paths of GDP per capita and other constraints. Here, these assumptions exist in either form
for CO2 intensity, exergy, the relationship between GDP per capita, the conditionalities for grant
receipt, and interest rates on debt repayments. Further, the inclusion of only two measures of social
protection expansion are considered. This was done for greatest reach with relatively easy
expenditure calculations. However, there is an argument to include broader social protection
measures, National Health Insurance being one example, with Annecke and Wolpe (2022) giving
others, that would affect the results. Secondly, some of the relationships between variables are
59
shown simply, without a more necessary in-depth analysis of their determinants. Whilst this would
be ideal to incorporate, it is beyond the scope of this paper. The incorporation of these
relationships, however, would be advantageous in an attempt to provide more concrete results and
constraints. In this paper, we aim to provide a basic framework of the dynamics of these
relationships and how they affect the various constraints towards achieving multiple objectives in
the environmental, social, and economic spheres. What the analysis could benefit from further is
the inclusion of feedback loops. In some instances, equations are mathematically linked, such that
a result from a constraint equation ‘further down the line’ may impact an assumption made earlier.
These cases are numerous, and this paper is limited by not including all of those dynamics. A
further limitation is the starting assumption: that South Africa will accommodate its negative
emission growth trajectory, with all other constraints being inline to accommodate that end. In
reality, this is highly unlikely to be the case, especially in an unchanging (or slowly changing)
energy generation and production structure. What this paper could benefit from is to start with a
given GDP per capita growth rate, a measure that has traditionally been more reliably estimated.
Regardless, if this were the case, the dynamics of the rest of the relationships would need to be
analysed, which has been carried out above. Further, the above analysis and its results are
conducted and presented in aggregate economic terms. We know however that this type of analysis
may be misleading due to the varying responsibilities of and impacts on different sectors. The
mining sector, for example, displays different historical dynamics than the retail trade sector. In
light of the scenario for negative emission growth, sectors such as agriculture may not be impacted,
though the utilities and chemicals sectors are likely to be hard-hit. A sector-level analysis and the
dynamics between direct and indirect emission and employment effects would constitute a stronger
paper. Further, the results here mention but do not provide indicative estimates for changes other
than if each constraint is met at its maximum. For example, a positive GDP per capita constraint
that leads to a positive employment growth constraint may not actually be met for employment,
having downstream effects. Of course, to include this would require a myriad of results beyond
what is useful. This paper provides mechanisms by which various constraints are technically,
mathematically implied, and that they can be achieved. Whilst mention has been made of the
constraints of political economy and developmental targets, no thresholds have been set to
incorporate this into the results of the paper. Whilst this would be contentious, the inclusion of
such thresholds may constitute more realistic results. For example, the role of strong trade
unionisation levels in the mining and minerals sector could play a huge role in transition efforts
(Rosenberg et al., 2022). There is also the omission of consideration for wider, more specific
though smaller in magnitude, areas of emission reduction. For example, possible avenues for
emission reduction via changing structures of modes of transport and the effects that greater public
transport infrastructure would have would be a welcome addition. Lastly, any analysis of the South
African economy is incomplete without a detailed consideration of its extremely high income,
wealth, spatial, and gender inequality. This, combined with the omission of effects that cash
transfers have on poverty and inequality (Goldman et al., 2020), constitute a major limitation of
the analysis.
In terms of further work, there are a few noteworthy extensions of the above analysis. Firstly, a
sectoral-level analysis would be of use in order to garner more precise policy implications useful
for policymakers. Ideally, this would include the more detailed and rigorous analyses between the
60
different constraints and their multiple dynamics, incorporating the feedback loops mentioned
above. Further, the inclusion of greater social security measures beyond that of cash transfers
would be of use. The effects of various revenue collection introductions and amendments in light
of this would be useful as well, especially in terms of the public debt constraint perspective. If
possible, a more in depth discussion on the implications of a balance of payments constraint and
its dynamics in tandem with a sectoral-level analysis would be a valuable contribution. Finally,
the manner in which these results and scenarios can be presented in a simple and efficient manner
would be an important contribution. With many processes and dynamics to consider,
communicating these results most effectively is something to consider going forward. To this end,
an interactive dashboard application would be useful so that there may be some public engagement
with the implications that these various assumptions and constraints have.
7. Conclusion
South Africa faces major hurdles in meeting targets in emission reduction, employment gains, and
social protection coverage. The period after the end of the commodity supercycle and the Great
Financial Crisis has brought precious few gains in these area, where there ought to have been in
line with these goals. COVID-19 has, in general, worsened these metrics further. Being the largest
carbon emitter in Africa and 12th globally, with an entrenched and carbon intensive energy
generation structure, there should be sufficient impetus to decarbonise the economy. This faces
political constraints, as well as financing considerations. With no protection for the working age
population, this will need to be addressed. Employment levels are also at a breaking point, with
the highest recorded unemployment level in history being recorded in the last quarter of 2021.
Assessing how these goals are to be met in the face of diverging results vs target metrics is of
utmost importance. Whether South Africa can achieve its NDC targets in this context depends on
concerted efforts of collaboration and consultation from all actors.
This paper provides a historical context in which these problems exist, leveraging a wide variety
of publicly available data to illustrate the evolution of these problems. From this, we are able to
analyse what recent trends have been and what, if any, events could be seen as turning points. This
historical analysis also allows for a broad assessment of the efficacy of previous and current policy
stances. As evident, policy changes have done little to improve trends in emission reduction, social
protection coverage, as well as employment gains. Further, this paper introduces a framework
within which to assess what some possible futures of South Africa hold. Starting with emission
reduction growth targets in line with those necessary to meet the upper and lower-bound targets of
the NDC, associated constraints are posited considering 4 main scenarios: no technological gains
and the upper-bound NDC target, no technological gains and the lower-bound NDC target, and
technological gains and the lower and upper-bound NDC targets.
For the most stringent case of no technological gains and emission growth paths in line with
meeting the lower-bound NDC target, the associated negative GDP per capita growth, negative
employment growth, increasing demand for social protection, upward pressure on the debt-to-GDP
61
ratio, and pressure from the balance of payment paint a dire picture. The elegance of the postured
methodological framework, however, is that it allows for changes to key input dynamics to be
altered in order to escape or at least arrest this downward spiral. On the upside, a scenario of
emission reductions with positive GDP growth leads to opportunities for South Africa to meets its
key targets without overbearing economic constraints. The likelihood of this scenario is
questionable, however, given frictions between actors, evidence of historical dynamics, and
financing requirements. In reality, this optimistic scenario would require an unparalleled effort of
coordination, efficiency, and timelines. Further, there are political constraints to consider, which
may prove to be the largest hurdles in this process. What is technically possible by means of the
analysis in this paper is not necessarily politically viable. A new social, environment-aligned social
compact with wide buy-in is required. At this moment, this seems distant.
Finally, what this paper aims to illustrate is that there are opportunities that lie in a transition away
from coal dependence in South Africa. Lowering emissions provides the chance to fundamentally
restructure an economy that has been spluttering for the last decade with little sign of improving.
The concept of strong sustainability, one that centres embeddedness and interconnected outcomes
of the environment, social sphere, and economy in light of limited capital substitutability should
be taken seriously, as it has been here, in order to meaningfully deconstruct the dynamics between
different policy objectives. It is through this lens that one may garner a holistic perspective of how
to approach a just transition, one that is mutually beneficial for all South Africans and that which
hosts them. Echoing popular civil society in recent years – the question is not whether we can
afford to do it, it is whether we can afford not to. It is the hope of this author that this paper provides
some insight as to how to think about the dynamics and constraints that come with moving towards
a future more fit for human and natural life.
62
Reference List
Annecke, W., Wolpe, P., 2022. What role for social policies in the framework of the just transition in
South Africa? (No. 230), Research paper. Agence Française de Développement.
Arndt, C., Davies, R., Makrelov, K., Thurlow, J., 2013. Measuring the Carbon Intensity of the South
African Economy. South African Journal of Economics 81, 393–415.
https://doi.org/10.1111/j.1813-6982.2012.01324.x
Ashman, S., 2021. SA’s climate crisis is embedded in coal and exports. New Frame. URL
https://www.newframe.com/sas-climate-crisis-is-embedded-in-coal-and-exports/
Atkinson, G., Dubourg, R., Hamilton, K., Munasinghe, M., Pearce, D., Young, C.E., 1997. Measuring
Sustainable Development. Edward Elgar Publishing.
Bhering, G., Serrano, F., Freitas, F., 2019. Thirlwall’s law, external debt sustainability, and the balance-of-
payments-constrained level and growth rates of output. Review of Keynesian Economics 7, 486–
497.
Bonini, A., 2018. Making data work for human development (Occasional Paper). UNDP Human
Development Report Office (HDRO).
British Petroleum, 2021. Statistical Review of World Energy, 2021. (database).
Burkhardt, P., 2021. Eskom Seeks Cheap Funding For $27 Billion Energy Plan. Bloomberg.com.
Business Tech, 2021. How many South Africans now rely on social grants: 1996 vs 2020. Available:
https://businesstech.co.za/news/government/459186/how-many-south-africans-now-rely-on-
social-grants-1996-vs-2020/
C19 People’s Coalition, 2021. OPEN LETTER: Government must extend the Covid-19 SRD grant! C19
People’s Coalition. Available: https://c19peoplescoalition.org.za/open-letter-government-must-
extend-the-covid-19-srd-grant/
CAIT, 2022. Greenhouse Gas (GHG) Emissions | Climate Watch | South Africa | Historical GHG emissions.
Available: https://www.climatewatchdata.org/ghg-emissions?end_year=2018&gases=all-
ghg®ions=ZAF§ors=total-including-lucf&source=CAIT&start_year=1990
Chandrasekhar, C.P., Ghosh, J., 2022. Who Controls Renewable Energy Technology? International
Development Economics Associates. Available: https://www.networkideas.org/featured-
articles/2022/05/who-controls-renewable-energy-technology/ (accessed 5.23.22).
ClimateLinks, 2018. Greenhouse Gas Emissions Factsheet: Democratic Republic of the Congo. Available:
https://www.climatelinks.org/resources/greenhouse-gas-emissions-factsheet-democratic-
republic-congo
Comrie, S., 2022. South Africa is pursuing major gas deal — and Russia wants in. Daily Maverick.
Available: https://www.dailymaverick.co.za/article/2022-03-27-south-africa-pursuing-major-
gas-deal-and-russia-wants-in/
COP26, 2021. Political Declaration on the Just Energy Transition in South Africa. UN Climate Change
Conference (COP26) at the SEC – Glasgow 2021. Available: https://ukcop26.org/political-
declaration-on-the-just-energy-transition-in-south-africa/
Costella, C., McCord, A., van Aalst, M., Holmes, R., Ammoun, J., Barca, V., 2021. Social protection and
climate change: scaling up ambition. SPACE.
Department of Energy, 2019. Integrated Resource Plan (IRP2019).
Department of Mineral Resources and Energy, 2022. Energy Sources: Coal. Available:
http://www.energy.gov.za/files/coal_frame.html
Department of Social Development, 2021. Rapid Assessment of the Implementation and Utilisation of
the Special Covid-19 Social Relief of Distress Grant. Available:
63
https://www.dsd.gov.za/index.php/latest-news/21-latest-news/410-the-rapid-assessment-of-
the-implementation-and-utilisation-of-the-special-covid-19-social-relief-of-distress-grant
Department of Social Development, 2020. Remarks by the Minister of Social Development, Ms Lindiwe
Zulu, on COVID-19 economic and social measures that were announced by President Cyril
Ramaphosa... Available: https://www.dsd.gov.za/index.php/latest-news/21-latest-news/127-
remarks-by-the-minister-of-social-development-ms-lindiwe-zulu-on-covid-19-economic-and-
social-measures-that-were-announced-by-president-cyril-ramaphosa
Ekins, P., Milligan, B., Usubiaga-Liano, A., 2019. A single indicator of strong sustainability for
development: Theoretical basis and practical implementation (No. 2019-112, Revised Draft, 21
December 2019), AFD Research Papers. Agence Française de Développement.
Enerdata, 2021. World Energy & Climate Statitics Yearbook 2021 (database).
Espagne, E., Godin, A., Magacho, G., Mantes, A., Yilmaz, D., 2021. Developing Countries’ Macroeconomic
Exposure to the Low-carbon Transition (Working Paper No. 220). Agence Française de
Développement.
Faku, D., 2020. R112bn and R16.4bn for Eskom, SAA in R128.4bn bailouts. IOL Online. Available:
https://www.iol.co.za/business-report/companies/r112bn-and-r164bn-for-eskom-saa-in-
r1284bn-bailouts-43548335
Farrington, J., Slater, R., 2006. Introduction: Cash Transfers: Panacea for Poverty Reduction or Money
Down the Drain? Development Policy Review 24, 499–511. https://doi.org/10.1111/j.1467-
7679.2006.00344.x
Godin, A., 2021. Strong sustainability in the long-run: squaring the circle?
Godin, A., David, A., Lecuyer, O., Leyronas, S., 2022. A strong sustainability approach to development
trajectories. European Journal of Economics and Economic Policy (forthcoming).
Godin, A., Hadji-Lazaro, P., 2020. Demand-induced transition risks: A systemic approach applied to South
Africa (No. 199), Research paper. Agence Française de Développement.
Goldman, M., Woolard, I., Jellema, J., 2020. The Impact of Taxes and Transfers on Poverty and Income
Distribution in South Africa 2014/2015. (No. 198), Agence Francaise de Developpement.
Hartley, F., Okunlola, A., Jacobs, D., Ntuli, N., Fourie, R., Borbonus, S., Nagel, L., Helgenberger, S., Burton,
S., Cunliffe, J., McCall, B., Caetano, T., Chiloane, L., 2019. Future skills and job creation through
renewable energy in South Africa: Assessing the co-benefits of decarbonising the power sector.
Hermanus, L., Montmasson-Clair, G., 2021. Making sense of jobs in South Africa’s just energy transition:
Managing the impact of a coal transition on employment. TIPS; Adapt.
Human, B.L., 2022. Finance Minister announces 2022 social grant increases. GroundUp News. Available:
http://www.groundup.org.za/article/finance-minister-announces-2022-social-grant-increases/
Huxman, M., Anwar, M., Nelson, D., 2019. Understanding the impact of a low carbon transition on South
Africa. Climate Policy Initiative.
IEA, 2021. World Energy Balances (database).
IEA, 2020. South Africa cumulative investment needs by scenario, 2019-2040 – Charts – Data &
Statistics. IEA. Available: https://www.iea.org/data-and-statistics/charts/south-africa-
cumulative-investment-needs-by-scenario-2019-2040
IEJ, 2021. Financing Options for a Universal Basic Income Guarantee in South Africa. Institute For
Economic Justice. Available: https://www.iej.org.za/financing-options-for-a-universal-basic-
income-guarantee-in-south-africa/
Institute for Economic Justice, 2021. Introducing a Universal Basic Income Guarantee for South Africa.
Institute For Economic Justice. Available: https://www.iej.org.za/introducing-a-universal-basic-
income-guarantee-for-south-africa/
International Labour Organisation, 2022. Statistics on social protection: population covered by social
protection floors/systems. ILOSTAT. Available: https://ilostat.ilo.org/topics/social-protection/
64
Isaacs, G., 2020. It’s not about debt, it’s about stimulating an economy in crisis. BusinessLIVE. Available:
https://www.businesslive.co.za/bd/opinion/2020-07-09-its-not-about-debt-its-about-
stimulating-an-economy-in-crisis/
Jain, R., Bassier, I., Budlender, J., Zizzamia, R., 2020. The labour market and poverty impacts of COVID-19
in South Africa: An update with NIDS-CRAM Wave 2.
Lund Committee, 1997. Report of the Lund Committee on Child and Family Support.
Mafata, M., 2022. Activists lobby Ramaphosa to make Covid-19 grant permanent. Fin24. Available:
https://www.news24.com/fin24/economy/activists-lobby-ramaphosa-to-make-covid-19-grant-
permanent-20220126
Makgetla, N., Maseko, N., Montmasson-Clair, G., Patel, M., 2019. National Employment Vulnerability
Assessment: Analysis of potential climate-change related impacts and vulnerable groups. TIPS.
National Planning Commission, 2020. Economic Progress Towards the National Development Plan’s
Vision 2030: Recommendations for Course Correction.
National Planning Commission, 2013. National Development Plan 2030: Our Future - make it work.
National Treasury, 2022a. National Budget - 2022 - Budget Data and Time Series in Excel Format - Table
6 (dataset).
National Treasury, 2022b. Budget 2022 Budget Review.
NDC Partnership, 2021. South Africa Takes The High Ambition Path | NDC Partnership. Available:
https://ndcpartnership.org/news/south-africa-takes-high-ambition-path
Our World in Data (OWID), 2022. Data on CO2 and Greenhouse Gas Emissions (database).
Pineda, A., 2021. The impact of renewable energy investment in South Africa: An exercise in Input-
Output analysis. EPOG Master Thesis.
Pradhan, P., Costa, L., Rybski, D., Lucht, W., Kropp, J.P., 2017. A Systematic Study of Sustainable
Development Goal (SDG) Interactions. Earth’s Future 5, 1169–1179.
https://doi.org/10.1002/2017EF000632
Presidential Climate Commission, 2021. Nationally Determined Contribution | Presidential Climate
Commission. PCC. Available: https://www.climatecommission.org.za/ndc (accessed 2.3.22).
Ramaphosa, C., 2022. State of the Nation Address (SoNA) by President Cyril Ramaphosa, Cape Town City
Hall, Thursday, 10 February 2022, 2022 State of the Nation Address.
Ramaphosa, C., 2021. From the desk of the President, Monday, 26 April 2021 | The Presidency.
Available: https://www.thepresidency.gov.za/from-the-desk-of-the-president/desk-
president%2C-monday%2C-26-april-2021
Ramaphosa, C., 2020. President Cyril Ramaphosa: South Africa’s response to Coronavirus COVID-19
pandemic | South African Government. Available: https://www.gov.za/speeches/president-cyril-
ramaphosa-south-africas-response-coronavirus-covid-19-pandemic-23-apr-2020
Reddy, T., Sokomani, A., 2008. Overview of the South African social security system, in: ISS Africa.
Republic of South Africa, 2021a. Minister Lindiwe Zulu: Re-introduction of special COVID-19 SRD grant |
South African Government. Speech. Available: https://www.gov.za/speeches/minister-lindiwe-
zulu-re-introduction-special-covid-19-srd-grant-4-aug-2021-0000
Republic of South Africa, 2021b. First Nationally Determined Contribution Under the Paris Agreement:
Updated September 2021.
Republic of South Africa, 1997. The Constitution of the Republic of South Africa.
Rigolini, J., 2021. Social Protection and Labor: A Key Enabler for Climate Change Adaptation and
Mitigation (Working Paper). World Bank, Washington, DC.
Rosenberg, J., Klassen, T., El khamlichi, S., 2022. Non-European perspectives on co-determination: the
cases of Canada and South Africa. https://doi.org/10.13140/RG.2.2.33844.53123
SAFSC, 2020. Towards a Shared Consensus on a #UBIGNOW. A Common Campaigning Position.
65
Schroder, E., Storm, S., 2020. Fiscal Policy in South Africa: Closed Input-Output Incomer and Employment
Multipliers.
Senona, E., 2020. Basic Income Support: a case for South Africa. Black Sash.
Sibembe, Y., 2019. How historical use of poor-quality coal has affected Eskom’s efficiency and brought
us load shedding. Daily Maverick. Available: https://www.dailymaverick.co.za/article/2019-03-
15-how-historical-use-of-poor-quality-coal-has-affected-eskoms-efficiency-and-brought-us-load-
shedding/
South African Social Security Agency, 2021a. Sixth Statistical Report: Payment System. Pretoria.
South African Social Security Agency, 2021b. Fifth Statistical Report: Payment System. Period: August
2021.
Spash, C.L., 2014. The Politics of Researching Carbon Trading in Australia (No. 2014/03), SRE-Discussion
Papers, SRE-Discussion Papers. WU Vienna University of Economics and Business.
Statista, 2022. Africa: CO2 emissions by country. Statista. Available:
https://www.statista.com/statistics/1268395/production-based-co2-emissions-in-africa-by-
country/
Statistics South Africa, 2022a. Quarterly Labour Force Survey 2021, Quarter 4.
https://doi.org/10.25828/KC6K-G071
Statistics South Africa, 2022b. GDP P0441 - GDP Time series 2021Q4.
Statistics South Africa, 2021a. Quarterly Labour Force Survey (QLFS), 3rd Quarter 2021.
Statistics South Africa, 2021b. National Poverty Lines 2021 (Statistical Release), P0310.1.
Stent, J., 2021. GroundUp Analysis: South African social grants — the research behind the controversy.
Daily Maverick. Available: https://www.dailymaverick.co.za/article/2021-11-11-south-african-
social-grants-the-research-behind-the-controversy/
Steyn, G., Tyler, E., 2021. What does South Africa’s Updated Nationally Determined Contribution Imply
for its coal fleet? (Briefing Note No. September (No. 2021/10)). Meridian Economics.
Steyn, L., 2021. Mantashe calls on Africa to unite against “coercion” by global anti-fossil fuel agenda.
Fin24. URL https://www.news24.com/fin24/economy/south-africa/mantashe-calls-on-africa-to-
unite-against-coercion-by-global-anti-fossil-fuel-agenda-20211109
Strauss, I., Isaacs, G., Rosenberg, J., Passoni, P., 2020. Rapid Country Assessment: South Africa - The
impacts from a COVID-19 shock to South Africa’s economy and labour market (Report).
Studies in Poverty and Inequality Institute, 2021. The Efficacy of Fiscal Policy in South Africa: Fiscal
Multipliers and the financing of a Basic Income Guarantee.
Taylor, V., 2020. Social protection & pathways to a basic income grant during & beyond COVID-19
(Presentation). National Planning Commission.
The Presidency, 2020. Statement by President Cyril Ramaphosa on further economic and social
measures in response to the COVID-19 epidemic. Available:
https://www.thepresidency.gov.za/speeches/statement-president-cyril-ramaphosa-further-
economic-and-social-measures-response-covid-19
Theys, J., Guimont, C., 2019. « Nous n’avons jamais été “soutenables” : pourquoi revisiter aujourd’hui la
notion de durabilité forte ? ». Développement durable et territoires. Économie, géographie,
politique, droit, sociologie. https://doi.org/10.4000/developpementdurable.13589
Turner, R., Pearce, D., Bateman, I., 1994. Environmental Economics: An Elementary Introduction.
Harvester Wheatsheaf, Hemel Hempstead.
UN Climate Change Conference UK 2021, 2021. Political Declaration on the Just Energy Transition in
South Africa. UN Climate Change Conference (COP26) at the SEC – Glasgow 2021. Available:
https://ukcop26.org/political-declaration-on-the-just-energy-transition-in-south-africa/
(accessed 2.3.22).
66
UN Department of Economic and Social Affairs, 2022. World Population Prospects - Population Division -
United Nations. Available: https://population.un.org/wpp/DataQuery/
UN General Assembly, 2015. Resolution adopted by the General Assembly on 25 September 2015.
United Nations, 1987. Our Common Future.
Van der Berg, S., 1994. Getting social security into gear: reflections on welfare reform in South Africa,
The Year 2000 International Research Conference on Social Security. Province of Eastern Cape
Department of Social Development.
van Seventer, D., Arndt, C., Davies, R.J., Gabriel, S., Harris, L., Robinson, S., 2021. Recovering from
COVID-19: Economic scenarios for South Africa. International Food Policy Research Institute,
Washington, DC. https://doi.org/10.2499/p15738coll2.134455
World Resources Institute, 2021. South Africa: Strong Foundations for a Just Transition.
67
Appendix
Figure A1. GHG emissions contribution by sector, 1990 – 2018.
Source: CAIT (2022)
Figure A2. GHG emissions, energy sector, 1990 – 2018.
Source: CAIT (2022)
68
Figure A3. Intensity of carbon emissions vs energy generation, 1990 – 2019.
Source: Authors construction. OWID (2022), International Energy Agency (IEA) (2021), Enerdata (2021), World
Development Indicators (WDI) (2022).
Note: CO2 measured in millions of tonnes, energy in petajoules.
Figure A4. Annual South African CO2 emissions per capita, 1900 – 2020.
Source: OWID (2022).
69
Figure A5. Primary energy consumption, South Africa, 1965 – 2019.
Source: OWID (2022).
Figure A6. GHG emissions, South Africa, 1990 – 2018.
Source: OWID (2022).
70
Figure A7. GHG emissions per capita, South Africa, 1990 – 2018.
Source: OWID (2022).
Figure A8. African GHG emissions per capita, 2018.
Source: OWID (2022)
71
Figure A9. CO2 and growth dynamics, 2022 – 2030.
Source: Authors construction.
Note: Exergy is constant at the 2019 level. Prefixes g_CO2 denote growth in CO2, g_CO2int denotes growth of CO2
intensity, g_pop is population growth, g_ycap denotes growth in GDP per capita. Suffixes _high and _low denote the
upper and lower-bound NDC targets respectively.
Figure A10. Adjusted CO2 and growth dynamics, 2022 – 2030.
Source: Authors construction.
Note: Energy increases at a rate of 5% per year. Exergy growth decreases linearly to 0.55PJ/ZAR billion by 2030.
Prefixes g_CO2 denote growth in CO2, g_CO2int denotes growth of CO2 intensity, g_pop is population growth,
g_ycap denotes growth in GDP per capita. Suffixes _high and _low denote the upper and lower-bound NDC targets
respectively.
