South A frican Green Economy Model (SAGEM)
Abstract
Green economy is a concept that covers several issues of sustainability. It is an economic
paradigm that prioritises increasing the well-being and equitable distribution of economic
benefits, while at the same time reducing environmental impacts. This paper introduces South
African Green Economy Model (SAGEM) that was developed to test the effects of investing
in green economy for selected sectors based on system dynamics approach. While the model
consists of 14 sectors and 31 modules, emphasis for green economy was on four key sectors
namely: natural resource management, agriculture, transport and energy. The baseline
simulation (2011 — 2030) and historical trends for specific variables over the period 2001 to
2010 is also presented.
Keywords: Green Economy, Sustainability, System dynamics, South Africa
1 Introduction
In recent years, there have been calls for new sustainable development pathways. This is due
to multiple and interrelated economic and environmental crises facing both developed and
developing countries. Several challenges derailing sustainable path include: energy crisis;
food crisis; water scarcity; biodiversity and ecosystem loss; climate security; desertification
(Gouvea et al., 2012). Green economy is gaining interest because it is considered to provide
potential to address these concerns (Cai et al., 2011; Xiaowei et al., 2011; Gouvea et al.,
2012).
Despite these interests, there are still limited studies within the scientific domain (peer
reviewed articles / studies) that models issues relating to green economy. Some these studies
include Gouvea et al.,(2012), who utilises green quadruple helix framework to evaluate how
water-intensive nations can develop additional competitive advantages in a green economy.
Carfi and Schiliro (2012) utilises a competitive model for green economy to address climate
change policy and creation of low-carbon technologies. Cai et al., (2011) utilises analytical
and input-output models to investigate the relationship between the green economy and green
jobs. These approaches however do not account for the cross-sectoral and dynamic nature of
green economy concept. Incorporating the green economy concept into a formal model
requires the use of mathematical model that integrates economic and physical dimensions of
the social, economic and environmental systems being analysed (Musango et al., 2012;
Department of Environmental A ffairs (DEA). and United Nations Environmental Programme
(UNEP). 2013 )
United Nations Environmental Programme (2011) utilised system dynamics to evaluate the
global green economy and accounts for the dynamic nature of the concept. System dynamics
uses stocks and flows to represent the system being investigated and is well suited to jointly
represent the economic, social, and environmental aspects of the development process. The
approach was developed at the Massachusetts Institute of Technology and has greatly
evolved over the last 60 years (see Forrester 1961; and Forrester 2007 for early and current
examples on the use of this methodology). In system dynamics models, causal relationships
are analysed, verified and formalised into models of differential equations (Barlas, 1996;
Sterman, 2000), and their behaviour is simulated. The approach is useful to analyse a variety
of development issues (Saeed, 1992; Saeed, 1998), including national policy analysis
(Pedercini and Barney, 2010).
While United Nations Environmental Programme (2011) provides a useful analysis, one
major criticism for the study is the lack of differentiation in relation to social equity and
economic context (Victor and Jackson, 2012). Thus, using the same approach as United
Nations Environmental Programme (2011), this paper presents a green economy model
developed for South Africa (here in referred to as South African Green Economy Model
(SAGEM). The purpose of the model was examine the question of whether equal or higher
growth could be attained with a more sustainable, equitable and resilient economy in which
natural resources would be preserved through more efficient use. The hypothesis was that,
correct management of natural resources does not necessarily imply accepting lower
economic growth going forward.
The development of the model was made possible through a partnership between the
Department of Environmental Affairs in South Africa and UNEP. Within this context, the
model was developed to explore the green economy transition for South Africa, with special
attention given to the ability to meet low carbon growth, resource efficiency, and pro-job
development targets.
2 Themodel (SAGEM)
The modelling process began with a workshop which was held late 2011 to conceptualize and
identify the needs for the green economy modelling for South Africa. During this meeting,
eight sectors were identified to have the potential to contribute to the green economy in South
Africa. These are: energy; agriculture; manufacturing; recycling (waste and management);
tourism; transport; water; ecosystem services (natural resource management).
In order to refine the sectors for the modelling, a Technical Stakeholder meeting was held in
February 2012. The primary objective of the workshop was to prioritize: (i) the sectors that
were to be focused; (ii) the targets to be aimed for; (iii) scenarios to be considered in the
modelling effort.
Due to time and data availability constraints, the model was to focus on four sectors in
analyzing green economy investment. The four main sectors that were selected include:
natural resource management; agriculture; transport and energy. The details for targets and
scenarios are found elsewhere (Musango et al., 2012; Department of Environmental Affairs
(DEA). and United Nations Environmental Programme (UNEP). 2013)
SAGEM utilised system dynamics approach following the T21 framework, which is a
planning tool that integrates the economic, social, and environmental dimensions of a country
into a single, comprehensive, transparent, user-friendly analytical framework. The model was
developed in Vensim software platform. In a broad sense, SAGEM was divided into fourteen
sectors (see Table 1) and 31 modules (see Appendix 1). A description of each of the sectors,
and their respective modules for SAGEM are discussed below.
Table 1: Green economy spheres and sectors
Natural resource Population Production
“management
Land Education Investment and
households
Water Health Government
Energy Employment
Emissions Public
“Minerals
2.1 Environment sphere
The environmental sphere of the SAGEM consists of 6 sectors and 18 modules (see Table 1
and Appendix 1) which are discussed below.
2.1.1 Natural resource management
The natural resource management represents the environmental and biodiversity protection
programmes, with a specific focus on the Working for Water programme. The sector is
classified into two modules. The first one calculates the water quantity provision with the
working for water programme, and consists of three stocks namely: accumulated restored
land (ARL ) ; cost of clearing invasive alien species (C ,,, ) ; and operating cost of maintaining
lap
restored land (oc ,,,). The accumulated restored land is represented as:
‘ap
ARL (t) = ARL (0)+ flr Jat
RT
Where, r,, is the restoration rate.
The second module estimates the potential electricity generation from invasive species. While
the primary objective of Working for Water is reducing the area of land under invasive alien
species, it has plans for value add activities. This is particular, generating electricity from
invasive alien species biomass. This is the key value add activity that was investigated in this
module, and it consists of one stock, namely, biomass plant capacity(Bpc ). This is
mathematically represented as:
BPC (t) =BPC (0)+ [Tyco - Taco ]At
Where, r,,.. and r,,, are biomass capacity construction rate and biomass capacity
depreciation respectively.
2.1.2. Land
This module represents the land use in South Africa and includes forest land (FL), crop
land(cL ), agricultural land (AL), conservation land(coL ), invasive alien species land
(IAL ) and other land (oL ). The invasive alien species is converted to other land and is
represented as:
IAL (t) = IAP (0) + [[r,, —ry Jat
7
Where, r,, and r,, are rate of spread from other land and rate of restoration to other land
sal
respectively.
Other land (01 ) can be converted to the different competing uses, and it is represented as:
OL (t) = OL (0)+ (LS (t+ Baa Pan DE or Tas Fan Tae Tar + Foxy) Jet
where: r,, is the rate of conversion from other land to settlement land; 1,, is the rate of
ol ‘ol
conversion from livestock land to other land (in ha/year); r,, is the rate of conversion from
oll
other land to livestock; r, is the rate of conversion from other land to conservation land;
r,, is the conversion rate from other land to forest land; r,, is the conversion rate form
lot ‘let
other land to crop land; and r,,, is the conversion rate from crop land to other land.
Although the land uses changes for the different land types the total land size in the country is
(obviously) maintained.
2.1.3 Water
The water sector consists of two modules: the water supply and water demand. The water
supply represents the yearly total water supply from renewable resources (TRWR ) and it is
utilised to estimate the water stress index(wsi ) , which influences the production sectors.
Total water demand from production sectors and domestic and municipal demand (Twp )
for water are also estimated in this module. For the case of water demand, specific attention is
given to the water demand requirement for the electricity generation sectors including: coal,
wind, solar, nuclear and biomass. The water stress index is therefore represented as:
2.1.4 Energy
The energy sector is categorised into energy production and energy demand. Energy
production consists of electricity supply, which is further categorised into coal, nuclear, wind,
pumped storage, hydro and solar; electricity technology generation share; and electricity
prices. The electricity supply from coal represents the capacity of coal electricity plant, the
amount generated given the capacity factor, and the potential for coal capacity reduction
given energy efficiency measures. It was assumed that the required electricity generation
from coal is the difference between the total electricity demand and the total electricity
generation from other electricity technologies. The coal electricity generation module
consists of two stocks, namely, coal energy capacity (cEc ) and potential cumulative coal
capacity reduction (PccR ) . These are represented as:
CEC (t) =CEC (0) + flr, =r, |dt
coe ~ Face
PCCR (t)=PCCR (0)+ f[r,,,, Jat
Where, r,,, is the coal plant construction rate; r,., is the rate of coal capacity depreciation;
r,,,, 1S the rate of potential electricity demand reduction.
pedr
Ina similar manner, other electricity supply modules, mainly hydro, nuclear, pumped storage,
solar and wind, represents the electricity generation from electricity technologies other than
coal. To illustrate with wind module, this consist of three stocks namely, wind plant under
construction (wec ) wind capacity (wc ) and decommissioned wind capacity (pwc ).
These are represented as:
WPC (t)=WPC (0)+{[r,. —r,. Jat
wpe Tue
WE (t) =WC (0)+ f[ ty. Toe Jat
DC (t) =DC (0)+ flr, Jat
‘awe
Where, r,,. is the rate of wind plant construction; r,, is the rate of wind capacity
completion; r,,, is the rate of decommissioning the wind capacity.
Electricity production is influenced by investments (installed capital capacity). The electricity
production is computed taking into account the demand and production capacities. Demand is
calculated by the sum of retail sales and distribution, distribution and transmission losses, and
the electricity net exports, which results in gross electricity demand. Subtracting the gross
electricity demand from the electricity generation from renewables, nuclear, hydro and
pumped storage yields the coal electricity demand for electricity production.
The technology share module estimates the proportion in which each electricity generation
technology contributes to the total electricity supply. In the case of electricity prices, this
module describes the electricity prices, which are taken as exogenous. This assumption is
7
reasonable for South Africa because the electricity prices are regulated by the National
Energy Regulator South Africa (NERSA). The electricity prices are projected exogenously
based on assumptions for NERSA’s determination on different electricity growth after 2013.
These are considered as: (i) BAU — 10%; (ii) average growth — 5%; and (iii) slow growth —
2.5%. Relative electricity prices have a major influence on production sectors, which in turn
influence GDP and investments.
Energy demand on the other hand consists of electricity demand, oil demand and gas demand.
These modules represent the drivers of energy demand in the medium- and long-term. The
electricity demand estimates the future electricity dynamics by the different electricity users,
and is driven by GDP, population and electricity prices. Oil demand is influenced by an
exogenously determined oil price since South Africa imports approximately 64% of its oil
consumption requirements. Gas production and consumption still plays an insignificant role
in the South African energy market. The gas demand module is therefore assumed to be
influence only by the GDP. It should be noted that the energy sub-model is estimated using a
variety of endogenous inputs (e.g. GDP and population) and exogenous inputs (e.g. the case
of electricity price).
2.1.5 Emissions
This consists of the air emission module which estimates the CO. emissions from the
different sectors. These sectors include industry, categorised as electricity and non-electricity
industry, which include: transport, agriculture, residential and services CO2 emissions. The
annual CO2 emission is endogenously determined in the model. The module consists of one
stock, cumulative air emissions (CAE ). This is increased by the rate of annual CO?
emissions (r,,) and decreased by decomposition of air emissions(r,,,). This is
mathematically represented as:
CAE (t) =CAE (0)+([1y, Ty, Jat
2.1.6 Minerals
This module represents the main mining activities, mainly coal, gold and PGM, and tracks
the mineral reserves, both unproven and proven reserves. The resources side of each of these
8
activities consists of two stocks, namely, undiscovered reserves and proven reserves. As an
illustration, gold proven reserves (GPR ) are increased by the rate of gold discovery ( r,, )
and decreased by the rate of gold extraction (r,,) . On the other hand, undiscovered reserves
(UGR ) are decreased by the rate of discovery. These are represented as follows:
GPR (t) =GPR (0)+ [[ty —ry Jat
UGR (t) =UGR (0)-flr,. Jat
In addition, the module calculates the corresponding energy demand and employment
generated from these sectors.
2.2 Society sphere
The societal sphere of the SAGEM model consists of 5 sectors and 7 modules (see Table 1
and Appendix 1), which are discussed below.
2.2.1 Population
This represents the population of South Africa and was categorized according to sex (male
and female) and age cohorts. The module consists of one stock, population (Pp), whose
dynamics is depended on births ( r,) , deaths ( r, ) and net migration ( r,,,). This is given as:
P(t) =P(0)+f[x, +r, —1, Jat
The module is generally used to dynamically estimate the factors that influence populations
(fertility rate and birth rate) and the way in which population influences environmental, social
and economic indicators in other modules such as water demand, energy demand, and GDP.
As an illustration, a growing population results in an increase in the water demand, which in
tum increases the total water demanded. With an increasing total demand, the water stress
index also increases, implying a reduction in the water reserve margin relative to the demand.
9
A decreasing water stress index consequently negatively influences the production sectors
(agriculture, industry and services), which in turn influences the size of GDP. The GDP, and
in particular, per capita income, has an influence on the fertility rate and life expectancy,
which in tun determines the level of population in the country (Figure 1). The main output of
the sub-model is population, which was compared with the nationally available population
data.
Figure 1: Causal loop of effect of population on water demand
2.2.2. Education
The module represents the advancement of the population through the education system,
from the school going children (both primary and high school), to becoming a literate
population. The module is categorised according to the South African education system of 7
years in primary school and 5 year in high school respectively. The government expenditure
in education and per capital income is assumed as the main influences of entrance to school.
The module consists of three stocks: students (s), who are increased by entrance rate
(r,,) and decreased by completion rate ( r,.) ; young literate population (YLP ) , who increase
due to completion of the education system; adult literate population (ALP ), increased by the
rate at which the young literate population become adults ( r,, ) .
These are represented as:
S(t) =8(0)+ [[r,—r,, Jat
10
YER (t) =YLR (0) + f[r,, 1, Jat
ALR (t) = ALR (0) + f[,, Jat
Generally, the module is utilised to estimate the access to education and the level of literacy
rate. These in turn are utilised to estimate the broader socio-economic factors such as
availability of labour, population and the GDP among others.
2.2.3 Health
This module aims to represent the access of basic health care based on the goverment
expenditure on health. While access to health has influence on fertility and life expectancy,
the sparse data for this sector did not allow for the model to be linked to other modules.
2.2.4 Public infrastructure (access to roads and transport)
This sector is categorised into roads infrastructure and transport. In the roads module, the
process of road construction is estimated as influenced by government expenditure on
transport and communication and the unit cost of roads construction. This is aimed at
estimating the access to roads, which has an influence in the production sectors. The module
consists of three stocks: roads under construction (RUC ), which is increased by the rate
road construction (r,,) and decreased by completion of construction (r,,); functioning
roads (FR), increased by roads completion (r,,) and decreased by disruption of roads
(x, ) ; and cost of maintaining roads, which is influenced by changes in cost of maintenance
with an assumed exogenous cost growth. These are represented as:
RUC (t)=RUC (0)+f[r,, —r,. Jat
FR (t) =FR(0) +f [toc —Tq Jat
The transport module on the other hand estimates the volume of roads, air and rail transport,
which is categorised according to the goods and passenger transport. The travel volumes are
calculated by multiplying the initial value of 2001 to the effects of GDP and population. The
11
assumption is that, all other things being equal, the travel volumes increases as GDP and
population increases. The associated emissions, employment and energy use from these
transportation modes is also estimated. These are calculated based on the CO2 emissions,
employment and energy consumption factor for each of the transport modes, which are
exogenously determined.
2.2.5 Employment
This represents the employment created in all the economic activities. The accumulation of
capital in the main production sectors (agriculture, industry and services) is considered
important in driving the growth in employment. The employment from production sectors
consists of three stocks: agriculture employment (AE) ; usual industry employment (UIE ) ;
usual services employment (USE ) . These stocks are influenced by the rate of net agricultural
hiring (r,,), net industry hiring( r,); and net services hiring( r,,) respectively, and are
represented as:
AE (t) = AE (0) +f [t,, Jat
UIE (t) =UIE (0) + [[x,, Jat
USE (t) =USE (0)+ f[r,, Jat
Estimation of employment from other sectors was disaggregated. These include: employment
from restoration natural resource management, with specific focus on the Working for Water
programme, employment from mining, transport and power sectors. The employment in all
the sectors tends to adjust over time to the demand. Therefore, employment cannot be more
than the supply of labour force.
2.3 Economy sphere
The economic sphere of the SAGEM consists of 3 sectors and 6 modules (see Table 1 and
Appendix 1), which are discussed below.
2.3.1 Production (agriculture, services, and industry)
This represents the agriculture, services, and industry sectors that are utilised to calculate the
gross domestic income.
12
The agriculture module module consists of one stock, the agriculture capital (ac ), which is
increased by rate of investment in agriculture (r,)and decreased by capital
depreciation( r,,, ) . This is represented as:
acd
AC (t)= AC (0) + f[1, - 1,5 Jat
The agriculture module also includes the crop production and differentiates between
production utilising conventional and organic fertilizer. The agricultural production is based
on the Cobb-Douglas production function, where land, labour and capital are the main factors
of production, and are influenced by water availability, electricity prices, literacy rate and
access to roads. Growth in the agriculture production is dependent in these factors of
production.
Being a production sector, it does have an influence on macroeconomic indicators related to
the green economy, as illustrated in Figure 2. An investment in ‘resource conservation’ and
‘agriculture capital’, will lead to an increase in ‘agricultural production’ with a consequent
increase in GDP with opportunities for further investments.
= __., Water stress
‘Agriculture
tabor
2
” Natural crop yield (efectve crop yea
SS AS
| * |
| | |
Organic | vA
fertizer_ | # |
(Teena)
“> fertilizer |
Harvested -
beat | ee,
‘ invest rm
x Y ai ~y
\ ¥ a peuture
mR Ohne = production
ea —___—
Figure 2: schematic representation of green economy effects of investment in agricultural
production
13
These investments may be in ‘health and education’, which will increase the ‘population’ and
‘labour force’, which can then also boost ‘agricultural production’; or it increases ‘education
levels’ that will improve ‘labour productivity’, and also ‘agricultural production’. Another
option is to channel investments into ‘pollution control’, which can improve ‘life expectancy’
and associate benefits to the rest of the economy, or directly improve ‘agricultural
production’. Many other casual loops are possible, for this, and other sectors.
Similarly, the industry and services modules represent the industry and services production
respectively, and employ Cobb-Douglas production function. Their production is also
influenced by water availability, electricity prices, access to roads and the education level.
The GDP module shows the accounting relationships in the calculation of the major income-
related indicators. These include the real GDP, which is influenced by the production sectors
and the per capita income among others. GDP is one of the main outputs of the economic
sector and the simulation results were similarly compared with the historically available data.
2.3.2. Investment and households
This represents the accounts of how the flows from the various economic production sectors
determine the investment in the country and the household income. The investment arises
from both private and public sectors, and is given as the sum of these investments from
various sources. The investment is then allocated to the various production sectors.
On the other hand, the household income is divided into consumption and savings. The
savings (PS) is a stock, which is increased by private savings (r,,) and decreased by
private investment( r,, e and eventually becomes part of the investment. This is represented
as:
Jat
Bi
Ps (t)=PS(0)+j[r, -r
14
2.3.3 Government
This module shows the various sources of government revenue including taxes, grants and
interest, which are received both domestically and from abroad. The module also shows the
government expenditure allocation to the various sectors.
2.3.4 Data collection
Based on the outcome of the workshop, the Sustainability Institute and Millennium Institute
collected data in consultation with, and assistance from, the Department of Environmental
Affairs, National Treasury, Department of Trade and Industry, and Development Bank of
Southem A frica, among others. The data that were collected covered the various sectors and
were obtained from various sources (see Table 2).
In utilizing the data, the approach was to first use nationally available data, or to use expert-
based documents in South Africa. In cases where these forms of information were not
available, the internationally best available data, such as the World Development Indicators
and data of the Intemational Energy Agency, were utilised. Where data were entirely
unavailable, assumptions were made based on the experience of the Millennium Institute
pertaining to the Green Economy Report (GER). The availability of data only allowed the
simulation to start in 2001, and the modelling period was set to 2030.
Table 2: Data sources for the SAGEM modules
IRONMENT
ater quantity provision with WiW Various documents from SA experts on Working for Water Programme
NRM — potential electricity generation from Various documents from SA experts on Working for Water Programme
invasive
Land STATS SA; World Bank Database (World Development Indicators); various documents on invasive
ien land
Water (demand and supply) Vater stress index
Water requirements in electricity generation ME / DoE Digest of Energy Statistics; Evans et al 2009
Electricity supply- coal STATS SA; DME / DoE Digest of Energy Statistics
lectricity supply- nuclear STATS SA; DME / DoE Digest of Energy Statistics
lectricity supply- hydro STATS SA; DME / DoE Digest of Eneruy Statistics
lectricity supply- pumped storage STATS SA; DME / DoE Digest of Energy Statistics
le energy ~ solar DME / DoE Digest of Energy Statistics; IRP 2010; SARi documents; Information on Engineering News
le energy — wind DME / DoE Digest of Energy Statistics; IRP 2010; SARi documents; Information on Engineering News
Electricity technology generation share STATS SA; DME / DoE Digest of Energy Statistics; IRP 2010
[Electiicity prices —SSSSSS«@DME/ DoE Digest of Energy Statistics; NERSA
Electricity demand STATS SA; DME / DoE Digest of Energy Statistics; World Bank Database (World Development
Indicators); International Energy Agency
Oil demand DME / DoE Digest of Energy Statistics;
Gas demand DME /DoE Digest of Energy Statistics;
Airemissions Intemational Energy Statistics; World Bank Database (World Development Indicators)
STATS SA Minerals statistics, World Bank Database (World Development Indicators)
SOCIETY
Population STATS SA; World Bank Database (World Development Indicators)
Education STATS SA; World Bank Indicators)
|Health (access to basic health) World Bank Database (World Indicators) |
45
{Roads (access to roads) Various South Africa documents; World Bank Database (World Indicators) ]
[Employment STATS SA; World Bank Database (World Indicators); Green jobs report |
Power employment SARi documents; Intemational Energy Agency
Agniculture STATS SA; World Bank Database (World Development Indicators)
Industry ; World Bank Database (World Tndicators)
Service ; World Bank Database (World Tndicators)
GDP ; World Bank Database (World Tndicators)
Investment and Household ; South African Reserve Bank
Government ; South African Reserve Bank
Transport DME / DoE Digest of Energy Statistics; World Bank Database (World Tndicators)
3 Model validation
According to Sterman (2000) validation is a continuous process of testing and building
confidence in the model. Models cannot be validated using a single test or ability to fit the
historical data. Thus, it is not generally possible or plausible to classify the model as correct
or incorrect (Sterman, 2000) but the model can be of good quality or poor quality (Barlas,
1996), suitable or not suitable. On a different note, Forrester (Forrester, 1961) argues that the
validity of system dynamics model cannot be discussed without reference to a specific
purpose. Thus, in order to make use of the standardized tests, it is always important to keep
note of the environment in which the model is designed to operate and the questions it aims
to answer. In short, validation enables one to: (i) understand whether the model is acceptable
for its intended use 47; and (ii) build confidence in the model based on the inferences of the
real system (Forrester, 1980; Barlas, 1996; Sterman, 2000).
Validation of SAGEM was an iterative process that ran throughout the modelling process.
Various validation tests were utilised and included:
direct structure validity test with the stakeholders and modelling team to ensure that
the model was consistent with the knowledge of the south African green economy
context;
parameter confirmation tests in situations where data was not available in south
African context in order to ensure it is consistent; and
behaviour tests to determine how the model output is consistent with the historical
data;
qualitative validation using expert opinion during workshops in order to improve
confidence in the usefulness of SAGEM.
16
4 Baseline simulation
The baseline simulation (BAU) of SAGEM is based on the assumption that the current trends
will continue The simulation replicates the historical trend over the period 2001 to 2010 and
assumes no fundamental changes in the policy or extemal conditions going forward to 2030.
This simulation was set up and calibrated to reflect the baseline projections for the various
existing sectoral models presented.
The real GDP is observed to grow over the simulation period (see Figure 3) due to the growth
of the production sectors, namely: services, industry and agriculture. The result of the GDP
simulation was compared with STATS SA data. For the past projections, the simulation
results performs well compared with the STATS SA data, with an R-square of 94.8% and an
average point-to-point standard deviation of 0.21%.
Between 2012 and 2030, the contribution for these production sectors to the GDP is expected
to increase by 68%, 14% and 44% for the industry, agriculture and services respectively.
Overall, this represents an increase in real GDP of 50.2% in 2030 relative to 2012.
Figure 3: Comparison of real GDP in BAU with data
The growth in the production sectors also correspond to the employment opportunities that
these sectors offer, with industry still providing much of the employment. The BAU shows a
24% and 36% increase in the persons employed in the services and industry sectors
We
respectively. On the other hand, agriculture shows a drop in employment of about 14% in
2030 relative to 2012.
The total employment covers all the sectors. This was disaggregated for the industry sector,
which also provides the employment in the mining and power sectors. Natural resource
management is also contributing to employment, resulting from the business as usual
allocation of some amount in the Working for Water programme. When considering the
employment creation of the specific sectors, the transport sectors show an increase in
employment by 2.3 times in 2030 relative to 2012 (see Figure 4).
Person m= NRM m Power sector m Transport @ Mining industry
600 000
Figure 4: Key sectors employment in the BAU scenario
The population similarly grows from 51.7 million in 2012 to 61.4 million in 2030 (see Figure
5). The simulation from 2001 to 2010 matches the historical data from STATS SA, with an
R-square of 97.9% and an average deviation of 0.09%. While the births are projected as
declining, the life expectancy is increasing hence reducing the deaths.
18
Figure 5: Comparison of population in BAU with data
Energy demand, that is electricity, oil and gas, are projected to grow, due to the growth in
both population and GDP. In 2030, these are projected to reach 121643 thousand TJ, 112783
TJ and 232644 GWh for oil, gas and electricity demand respectively. In the case of electricity
generation, the share of coal generated electricity remains relatively highly, though it declines
from 91.7% in 2012 to 83.6% in 2030. This is due to the introduction of the renewable energy
that is already committed in the BAU case. It should however be noted that the demand for
electricity is unmet in some years of the BAU case.
In a similar manner, water demand is also projected to increase due to the growth in
population and GDP. The demand reaches 13955 billion litres, representing 8% increase
relative to 2012. With South A frica being a water stress country, receiving only an average of
500mm rainfall per annum, an increase in the water demand obviously increases the water
stress index. To avoid compromising the already stressed water resources, water management
practices would therefore be necessary.
In terms of land use changes (see Figure 6), the cropland increases to 16.5 million hectares by
2030, representing a 9% growth relative to 2012. This expansion, though marginal, is due to
the increasing food demand from the growing population.
19
Ha {in m=Crop m Forest m Invasive species m Livestock
million)
Figure 6: Selected land use changes in the BAU scenario
The area infested by the invasive alien species is observed to increase by 28% in 2030
relative to 2012. While the working for water programme is incorporated in the BAU, the
rate of spread of invasive alien species is greater than the rate of restoration — hence the
increase in the invasive alien species land. With similar programmes and targets in the BAU
on the Working for Woodlands, the forestland is observed to increase by 0.9% per annum
from 2012 to 2030. Similarly, settlement land also grows by 1% on average, reaching 2.14
million hectares. On the other hand, the growth in livestock land, and area infested by
invasive alien species, decreases by 2% and 6% respectively.
The simulation result of the annual CO emissions fits well with the World Development
Indicators data, with an R-square of 80% and average deviation of 0.09% (see Figure 7). The
annual CO, emissions are observed to be relatively increasing, reaching 475 billion kilograms
by 2030. This is as a result of the increasing energy demand, population and GDP. The key
sectors contributing to the increasing emissions are the energy, mainly due to power
generation, and transport sectors (see Figure 8).
20
@ Transport m Power
Figure 8: BAU CO, emissions in transport and power sectors
The power generation sector contribution to CO2 emissions is projected to reach 297 million
tons. Initially, the emissions from the power generation are increasing. This is due to the
planned coal generation that was modelled in the BAU case. In addition, the renewable
energy generation that is committed in the IRP2010 was modelled as part of the BAU
scenario. This explains why the CO> emissions are growing at a decreasing rate. On the other
hand, the transport sector emissions are increasing, with a growth rate of 1.6% per annum,
between 2012 and 2030. The share of transport sector to CO» emissions is thus projected to
rise from 11.1% in 2012 to 13% in 2030.
21
5 Conclusion
This paper introduced the South Africa Green Economy Model (SAGEM) that was intended
to assess the impact of green economy investment of four key sectors. It was developed to
evaluate the impact of green economy investment on medium- to long-term environmental,
economic and social development issues. Given the data availability at a national scale, the
time horizon for the model begins in 2001 and extends to 2030, which is in line with the
current NDP time horizon. The simulation could also be easily extended further in the future
if need be. The historical trends from 2001 to 2010 were utilised to ensure that the model
replicates the characteristics of the behaviour of the issues investigated in SAGEM.
SAGEM does not capture all the inherent aspects that are necessary to allow for transitioning
to green economy. For instance, the model only represents the national environmental, social
and economic spheres without disaggregation at provincial or cities level. In addition,
SAGEM only represents the environmental, social and economic spheres at a country level,
without disaggregation at provincial or cities level. Additionally, SAGEM does not address
the sources of funding for green economy and the different agents that may be responsible for
transitioning towards green economy is not explicitly taken into account. Despite these
limitations, the key contribution of SAGEM is its dynamic nature, cross-sectoral analysis and
endogenous feedback loops within the various spheres, sectors and modules that are
considered towards achieving green economy objectives (job-creation, low carbon transition
and economic growth) in South Africa. The model has a potential to be extended to include
other sectors that were initially highlighted as having a potential to contribute to green
economy but were not modelled in detail. As an illustration, SAGEM is now currently being
extended to include the manufacturing sector, in order to understand how the sub-sectors such
as automobile, agro-processing and paper and pulp industry.
22
Appendix 1: All modules of SAGEM
Natural resource Management sector Population sector Production sector
1, Water quantity provision with W£W 19. Population 26. Agriculture
2. Potential electricity generation from invasive 27. Industry
28, Services
29. GDP
Land sector Education sector Households and investment sector
3. Land 20, Education 30a) Household accounts
30b) Banks
Water sector Health sector Government sector
4, Water demand and supply 21. Access to basic health 31, Govemment accounts
5, Water requi in electricity generation
Energy sector Employment sector
Energy production 22, Employment in different sectors
6. Electricity supply ~ coal = Industry employment
7. Electricity supply — nuclear - Agriculture employment
8. Electricity supply ~ hydro - Services employment
9, Electricity supply- pumped storage - Employment from NRM (invasive alien)
10. Electricity supply ~ solar 23, Power sector employment
11. Electricity supply ~ wind
12. Electricity technology generation share
13. Electricity prices
Energy demand
14, Electricity demand
15. Oil demand
16. Gas demand
Emissions Public infrastructure sector
17, Emissions from different sectors 24, Transport
-Power sector emissions 25. Access to roads
-Industy emissions
-Transport emissions
-Agriculture emissions
Minerals
18. Mining
“Coal
-Gold and uranium
23
REFERENCES
BARLAS, Y. (1996) Formal aspects of model validity and validation in system dynamics. System
Dynamics Review, 12, 183-210.
CAL, W., WANG, C., CHEN, J. & WANG, S. (2011) Green economy and green jobs: Myth or
reality? The case of China's power generation sector. Energy, 36, 5994-6003.
CARFI, D. & SCHILIRO, D. (2012) A coopetitive model for the green economy. Economic
Modelling, 29, 1215-1219.
DEPARTMENT OF ENVIRONMENTAL AFFAIRS (DEA). & UNITED NATIONS
ENVIRONMENTAL PROGRAMME (UNEP). (2013 ) Towards and integrated South
African Green Economy Model (SAGEM): focus on the sectors of Natural Resource
Management, Agriculture, Transport and Energy, (In Press).
FORRESTER, J. W. (1961) Industrial dynamics, Cambridge, MA, Productivity Press.
FORRESTER, J. W. (1980) System dynamics - future opportunities. INLEGASTO, A.A. &
FORRESTER, J. W. (Eds.) System dynamics. TIMS studies in the management sciencs Vol.
14. Oxford, North-Holland.
GOUVEA, R., KASSICIEH, S. & MONTOYA, M. J. R. (2012) Using the quadruple helix to design
strategies for the green economy. Technological Forecasting and Social Change.
MUSANGO, J. K., BRENT, A. C. & BASSI, A. M. (2012) Modelling the transition towards green
economy in South Africa. Paper submitted to Technology Forecasting and Social Change, 30
November 2012 [Manuscript Number: TFS-S-12-00585, Under Review].
PEDERCINI, M. & BARNEY, G. O. (2010) Dynamic analysis of interventions designed to achieve
Millennium Development Goals (MDG): The case of Ghana. Socio-Economic Planning
Sciences, 44, 89-99.
SAEED, K. (1992) slicing a complex problem for system dynamics modeling. System Dynamics
Review, 8, 251-262.
SAEED, K. (1998) Towards sustainable development: essays on system analysis of national policy,
Aldershot, UK, Ashgate Publishing Company.
STERMAN, J. D. (2000) Business dynamics: systems thinking and modelling for a complex world. ,
New Y ork, McGraw-Hill/Irwin.
UNITED NATIONS ENVIRONMENTAL PROGRAMME (UNEP). (2011) Towards a green
economy: pathways to sustainable development and poverty eradication. Nairobi: UNEP.
Available online:
http://www.unep.org/greeneconomy/greeneconomyreport/tabid/29846/default.aspx.
VICTOR, P. A. & JACKSON, T. (2012) A Commentary on UNEP's Green Economy Scenarios.
Ecological Economics, 77, 11-15.
XIAOWEI, J., QL, S. & YANFENG, G. (2011) New Approaches to the Green Economy of China in
the Multiple Crises. Energy Procedia, 5, 1365-1370.
24
