EFFECTS OF ECONOMIC GREENHOUSE GASES EMISSIONS
ON ECONOMIC
GROWTH: EVIDENCE FROM NIGERIA AND SOUTH AFRICA
BY
1Asogwa,
Ikechukwu.S.2Anumudu Charles, N. & 3Alugbuo Justin, C.
Department of Economics, Michael Okpara University of Agriculture, Umudike,Nigeria.
Department of Economics, University of Nigeria Nsukka.
Email: [email protected]
At
40th IAEE International Conference, Singapore
Outline

Overview

Problem Statement

Literature Review

Methodology

Result and Discussion

Recommendation and Conclusion
Greenhouse Gases Emission Intensity and
Sources
South Africa - 90
South Africa - 92
South Africa - 94
South Africa - 96
South Africa - 98
South Africa - 00
South Africa - 02
South Africa - 04
South Africa - 06
South Africa - 08
South Africa - 10
South Africa - 12
Nig eria - 91
Nig eria - 93
Nig eria - 95
Nig eria - 97
Nig eria - 99
Nig eria - 01
Nig eria - 03
Nig eria - 05
Nig eria - 07
Nig eria - 09
Nig eria - 11
1E+11
30
20
0E+00
900,000
GDP
25
800,000
700,000
20
600,000
500,000
400,000
300,000
5
200,000
0
200,000,000
POP
160,000,000
120,000,000
80,000,000
40,000,000
0
South Africa - 90
South Africa - 92
South Africa - 94
South Africa - 96
South Africa - 98
South Africa - 00
South Africa - 02
South Africa - 04
South Africa - 06
South Africa - 08
South Africa - 10
South Africa - 12
Nig eria - 91
Nig eria - 93
Nig eria - 95
Nig eria - 97
Nig eria - 99
Nig eria - 01
Nig eria - 03
Nig eria - 05
Nig eria - 07
Nig eria - 09
Nig eria - 11
AGRICULTURE
CAPITAL
40
6E+10
300
4E+10
250
INDUSTRIES
South Africa - 90
South Africa - 92
South Africa - 94
South Africa - 96
South Africa - 98
South Africa - 00
South Africa - 02
South Africa - 04
South Africa - 06
South Africa - 08
South Africa - 10
South Africa - 12
Nig eria - 91
Nig eria - 93
Nig eria - 95
Nig eria - 97
Nig eria - 99
Nig eria - 01
Nig eria - 03
Nig eria - 05
Nig eria - 07
Nig eria - 09
Nig eria - 11
50
South Africa - 90
South Africa - 92
South Africa - 94
South Africa - 96
South Africa - 98
South Africa - 00
South Africa - 02
South Africa - 04
South Africa - 06
South Africa - 08
South Africa - 10
South Africa - 12
Nig eria - 91
Nig eria - 93
Nig eria - 95
Nig eria - 97
Nig eria - 99
Nig eria - 01
Nig eria - 03
Nig eria - 05
Nig eria - 07
Nig eria - 09
Nig eria - 11
South Africa - 90
South Africa - 92
South Africa - 94
South Africa - 96
South Africa - 98
South Africa - 00
South Africa - 02
South Africa - 04
South Africa - 06
South Africa - 08
South Africa - 10
South Africa - 12
Nig eria - 91
Nig eria - 93
Nig eria - 95
Nig eria - 97
Nig eria - 99
Nig eria - 01
Nig eria - 03
Nig eria - 05
Nig eria - 07
Nig eria - 09
Nig eria - 11
60
South Africa - 90
South Africa - 92
South Africa - 94
South Africa - 96
South Africa - 98
South Africa - 00
South Africa - 02
South Africa - 04
South Africa - 06
South Africa - 08
South Africa - 10
South Africa - 12
Nig eria - 91
Nig eria - 93
Nig eria - 95
Nig eria - 97
Nig eria - 99
Nig eria - 01
Nig eria - 03
Nig eria - 05
Nig eria - 07
Nig eria - 09
Nig eria - 11
South Africa - 90
South Africa - 92
South Africa - 94
South Africa - 96
South Africa - 98
South Africa - 00
South Africa - 02
South Africa - 04
South Africa - 06
South Africa - 08
South Africa - 10
South Africa - 12
Nig eria - 91
Nig eria - 93
Nig eria - 95
Nig eria - 97
Nig eria - 99
Nig eria - 01
Nig eria - 03
Nig eria - 05
Nig eria - 07
Nig eria - 09
Nig eria - 11
Emission and Growth Outlook for Nigeria and
South Africa(1990-2012)
450
ENERGY
8E+10
400
350
2E+10
200
150
100
70
WASTE
60
15
50
10
40
30
20
10
Overview

Greenhouse Gases (GHG) emissions pose a high threat to our climate system as argued
in various literatures. The global warming of 2.5oC as obtainable is said to be caused by
Greenhouse Gases Emissions sourced from economic activities geared towards
economic growth. Greenhouse gases emissions cost could be referred to as the actual
cost and effects of the negative externality on the economy.

Externalities in its two sides of a coin can be “negative or positive” and virtually, a
resultant from economic activity just as externalities from Greenhouse gases emissions
cause varieties of social ill with substantial cost on the economy measured in health,
productivity and potential cost incurred by global climate change.
Overview(cont’)

Economic activities beget Economic Growth and the implications of uncontrolled
economic activities results to the indiscriminately emissions of gases into the atmosphere
that lead to the preponderance of varieties of social ills as summarized above as
“negative externalities”.

The target of every developing and developed economies is to achieve an improved
economic growth and economic development. The major economic goal suggests that to
grow economic growth that necessitates an aggressive economic transformation through
increasing share of industrial, agricultural, and manufacturing activities in the Gross
Domestic Product is not substitutable for a growing economy.
Problem Statement

Among selected Africa countries, Co2 emission in South Africa and Nigeria is
higher than other Africa countries with 462.60 and 296.68 respectively. This is
the emission in MT Co2 (excluding Land use Change and forestry) as reported by
(WRI 2015).

South Africa and Nigeria in their unique and common features both in
geographical location and endowments are concerned over climate change cause
by the economic activities which include: transportation, biofuel, combustion,
agriculture, oil exploration etc which contribute to the GHG emission rate.
Research Questions

What are the effects of greenhouse gases emission by sector
classification on Productivity in Africa?

Is the Effects of Emission caused by economic activities on African
Economic growth Significant”?
Literature Review

Rezai et al (2009) applied the Keynes-Ramsey growth model and argued that “it
is socially beneficial for the present and near future generation to scarify their
own consumption to mitigate global warming for the benefit of generation yet to
come.

Mendelsohn,(2009) argued in support that the long term consequences of climate
change have given the impression that the climate impact from greenhouse gases
threaten long term economic growth. Upon the growing greenhouse gases
emission, the determinants of the greenhouse gases emission are seen as the force
driving activities of economic growth.
Literature Review Cont’

The driving force activities include fossil energy demand or
consumption, rent from forestry trade, agricultural land area
expansion and farm technology as stated by (Achike and Onoja,
2014).

The aggregate production function links the total output of an
economy to the aggregate amount of labour, human capital and
physical capital in the economy and some simple measure of the level
of technology in the economy (Abhijit and Esther 2004).
Methodology

To model the effects of greenhouse gases emission on economic growth, we apply the theoretical
model of Solow growth model to explain for the various interactions of factors of production given
economic activities that lead to economic growth using a cross section econometric approach as in
Olarinde, Martin and Abdulsalam (2014).

The stability properties of the variables employed will be examined in order to determine and
evaluate the variables’ stochastic properties. Some of the stability properties that will be examined
include: Stationarity, Cointegration and diagnostic econometric test like: cross sectional
dependence test.

Data from World Resource Institute, Climate Analysis Indicator and Tool for the period of 1990 to
2014 across Nigeria and South Africa emission record
Methodology Cont’ : Model Building of the Study

The functional form of Solow Growth model with cobb Douglas extension as in
(Debasish 2013) is specified as below
Y(t) = K(t)α H(t) β (A(t) L(t))1−α −β ,...(III) 0<α+β<1

Where t denotes time. The Augmented Human capital in the Solow Model explains the
effects of human capital on growth process. The activities to economic growth and
greenhouse gases emissions are sourced from economic activities. Here, H is defined as
stock of human capital, β is the share of human capital in total output. The assumption of
0<α+β<1 indicates decreasing return to scale.
Methodology Cont’:Model functional Form

Following the literature and standard Solow growth model as above we relate the economic growth to
the key greenhouse gases emission from the source sector of economic activities, labour force and
Population

just
as
we
represent
in
its
functional
form
below:
GDP(t)/ LABOUR = F(Technological Progress, Capital-Labour ratio, Industries-GHG, AgriculturalGHG ,Waste-GHG, Energy-GHG ) ……………………………………………. . .(II)

Where Y(t) is the gross domestic product for Nigeria and South Africa, IGHG is Industries greenhouse
gases emission, MGHG is the manufacturing greenhouse gases emission, AGHG is the Agricultural
greenhouse gases emission, POP is the population proxy labour, EDU is the Augmented Human capital,
Capital.
Methodology Cont’: Pre Model Estimation Test: Panel Unit Root Test.

Panel unit root test are closely related to the unit root test carried out on a single
series but it cannot be considered as the same. In the year 1993, Levin and Lin
established the foundation for panel unit root test and thereafter, few more test
emerged such as: Im-Pesaran-Shin (1997) and Maddala (1999) as recorded by
(Hoang and Mcnown, accessed online: www.spot.colorando.edu, 2016).

The test methods adopted in this study claimed the presence of unit root
(common unit root) in the series as the null hypothesis while the alternative reads
the absence of common unit root.
Pre Model Estimation Test: Cross Section Unit Root Test. Cont’

H0:α = 0; H:α < 0

At 5% level of significance.

The Levin, Lin and Chu test statistics for unit root test as stated by is asymptotically
normally distributed as stated by eviews tutorial class accessed online,2016 is below
^
t − (NT )SNσ −2 Se(α )µmT
T=
α

→ N(0,1)......................
α
(V )
σmT
Where tα is the standard t statistics for estimated α = 0, σ2 is the estimated variance of the error
term η, Se(ά) is the standard error of estimated α and SN is the mean of the ratio of the Long run
standard deviation to the innovation standard deviation for each individual,µmt and σmt are the
adjustment terms for the mean and standard deviation.
The result of the panel unit root test is as below:*Denote the test statistics while **
Test
.
Log(Waste
Log(energy
Method
(Ind.GHG)
GHG)
GHG)
Log(output)
denotes the P.Value
of each test Log
Log(K-Lratio)
Log(Agriculture GHG)
Levin, Lin
-0.95630*
2.38659*
3.90192*
0.25277*
0.35503*
-1.0750*
& Chu t*
(0.1695)**
(0.9915)**
(1.0000)**
(0.5998)**
(0.6387)**
(0.1412)**
Breitung t-
0.65560*
2.78967*
-3.47225
-2.12687*
1.60332*
-1.61587*
Stat
(0.7440)**
(0.9974)**
(0.0003)
(0.0167)**
(0.9456)**
(0.0531)**
Im, Pesaran
-1.05007*
2.66750*
5.81817*
-0.79088*
-0.02894*
-0.56201*
and Shin
(0.1468)**
(0.9962)**
(1.0000)**
(0.2145)**
(0.4885)**
(0.2871)**
ADF -
7.18358*
2.05966*
0.000076*
5.34571*
3.34660*
4.76776*
Fisher Chi-
(0.1265)**
(0.7248)**
(1.0000)**
(0.2536)**
(0.5016)**
(0.3120)**
PP - Fisher
6.89351*
0.94245*
0.00441*
5.34571*
1.60057*
4.76776*
Chi-square
(0.1416)**
0.9184**
(1.0000)**
(0.2536)**
(0.8087)**
(0.3120)**
Remarks
Not
Not
Not
Not stationary@
Not
Not stationary@ levels
stationary@
stationary@
stationary@
levels
stationary@
levels
levels
levels
W-stat
square
levels
Correction of Panel Series Unit Root: observed below that all the
variable are stationary after the first difference except the log of emission from industries that shown
stationarity atL(output
second difference
Test
L(Ind.GHG)
Log(waste
Log(energy
Log(KLLog(Agriculture
L(Ind.GHG)
GHG)
GHG)
ratio)
GHG)
@2nd Diff
0.83732*
-0.22626**
-4.68196*
-3.57767*
-4.69353*
-2.28670*
(0.0001)**
(0.7988)**
(0.4105)**
(0.0000)**
(0.0002)**
(0.0000)**
(0.0111)**
Breitung
-1.89603*
1.33532*
-0.30252
-3.87655*
-1.79826*
-2.26849*
-1.26265*
T-stat
(0.0001)**
(0.9091)**
(0.3811)**
(0.00011)**
(0.0361)**
(0.00116)**
(0.1034)**
Im,
-1.95993*
-1.82454*
-2.51270*
-3.551180*
-2.22779*
-3.92331*
-3.52684*
Pesaran
(0.0250)**
(0.0340)**
(0.0060)
(0.0002)**
(0.0129)**
(0.0000)**
(0.0002)**
ADF -
10.1441*
9.54701*
13.2869*
17.4504*
11.7675*
20.4746*
17.3714*
Fisher
(0.0381)**
(0.0488)**
(0.0100)**
(0.0016)**
(0.0192)**
(0.0004)**
(0.00116)**
PP -
9.51859*
3.03828*
13.2869*
17.5413*
11.7694*
23.5639*
16.3546*
Fisher
(0.0494)**
0.5514**
(0.0100)**
(0.0015)**
(0.0192)**
(0.0001)**
(0.0026)**
Stationary
not
Stationary
Stationary @
Stationary
Stationary @ 1st
Stationary
Stationary @
@ 1st Diff
1st Diff
@ 1st Diff
Diff
@ 2nd Diff
Method
GHG)
Levin,
-3.73030*
Lin &
Chu t*
and Shin
W-stat
Chisquare
Chisquare
Remarks
@ 1st Diff
1st Diff
Model Specification and Estimation

Solow growth model on the long-run of output per worker depends on
technological progress .The determination of the short run growth source is a case
of empirical approach.
gdp (t )
capital (t )
  it  1it log(indus )   2it log( waste)   3it (energy )   4it log(
)
labour (t )
labour (t )
 5it log(agriculture)   6it A  Z it  it ...( II )
productivity  log

Where “indus” represents the emission from industrial economic activities, waste
is the emission from

any other sectors of the economy not captured in the class, energy is the emission
from energy consumptions and agriculture variable represent all the emission
from the agricultural activities. The technological progress is denoted by A in the
model and it will be captured by the deviation from the regression(Solow
residual).

The convergence assumption of the Solow growth model to a growth balance
path will be evaluated by the βit, if the βit is negative, it suggest that convergence
with higher initial incomes have lower growth of growth which is caused the
incentive for capital to flow from rich to poor nation
Result and Discussion
Cross section Weights (PCSE) standard error and covariance (no df correction)
Dependent Variable
(productivity) output ratio
Coefficient
Standard Error
t-Statistics
Prob
Constant
-0.832852
0.400609
-2.078965
0.0455*
D(LOG(K-LRATIO),1)
-0.048589
0.042691
-1.1381151
0.2633
Technological Progress(A)
0.094910
0.042532
2.231473
0.0326*
D(LOG(INDUSTRIES),2)
-0.077860
0.076636
-1.015970
0.3170
D(LOG(WASTE),1)
2.033103
1.452013
1.4001196
0.1708
D(LOG(ENERGY),1)
0.182996
0.170036
1.076215
0.2896
D(LOG(AGRICULTURE),1)
CROSS SECTION
-0.312370
0.278646
-1.121026
0.2704
EFFECT
NIGERIA
-0.017212
SOUTH
AFRICA
R-Square
0.287000
Mean
0.041986
dependent Var
Adjusted R-
0.135757
S.D dependent
0.046525
0.043252
Var
Sum Square
0.061733
1.698788
resid
F-statistics
1.897614
Prob(F-statistics)
0.101649
Akaike Info
-3.270387
Schwarz Crit
-2.936031
Hannan-Quinn
-3.148633
Squared
S.E of regression
Durbin-Watson
Stat
0.018072
Discussion cont’

From result Table, all greenhouse gases emission sourced from the various
economic activities across the countries (Nigeria and South Africa) has a
negative effect or poses negative externality on the effective labour
productivity in the countries except waste emission and emission from
energy sector

Waste emissions are considered as every other emissions of greenhouse
gases that may not have direct relationship with economic activities as in
the other sectors emission specified.
Discussion cont’

The Labour effective capital defined as capital-labour ratio suggest the
convergence assumption of the Solow growth model and the possibility of
balanced growth path in the region supported with the negative sign on the
effective output productivity 4.868 percent (-0.048589). The convergence speed
is also slow against the desired perfect (-1) convergence magnitude or close to
one convergence magnitude for perfect speed.

Therefore, the technological progress is positively related to the effective labour
productivity with a unit increase in the technological progress to lead to
9.4%(0.094910) increase in the effective labour productivity, all things being
Post Estimation Test
Cross Sectional Dependence Test

The common assumption that the disturbance in a cross section data
model is cross sectional independent is not always consistent in a cross
section model. Recent literature has it that Ignoring cross-sectional
dependence in estimation can have serious consequences, with
unaccounted for residual dependence resulting in estimator efficiency
loss and invalid test statistics. The some test of cross section
dependence include: Breusch-Pagan(1980) LM, Pesaran(2004) scaled LM,
Baltigi,Feng and Kao(2012) Bias corrected scaled LM and Pesaran (2004)
CD.
Post Estimation Test Cont’

Residual Cross section Dependence Test

Null hypothesis: No cross-section dependence (correlation) in residual.
Test
Statistics
d.f.
Prob
Breusch-Pagan LM
0.115661
1
0.7338
Pesaran Scaled LM
-2.03953
0.0414
-2.089536
0.0367
0.340089
0.7338
Bias-corrected
scaled
LM
Pesaran CD
Test
Why Cross section Weights (PCSE) standard error and
covariance (no df correction) in the result table?
Statistics
d.f.
Prob
Breusch-Pagan LM
0.115661
1
0.7338
Pesaran Scaled LM
-2.03953
0.0414
-2.089536
0.0367
0.340089
0.7338
Bias-corrected
scaled
LM
Pesaran CD
Cont’

From table above, the null of no cross sectional dependence could not be rejected given
the probability values of Breusch-Pagan LM (0.7338) and Pesaran CD (0.7338) at 5%
level of significance. This therefor suggest a cross sectional independence of the error.
Although this is in opposite of the probability values of Pesaran Scaled LM (0.0414) and
Bias-corrected scaled LM (0.0367). The balance in the test therefor suggest a correction
test of the cross sectional dependence such as Panel-Corrected Standard Error (PCSE).
The Panel Corrected Standard error (PCSE) estimation perform substantially better than
the asymptotically efficient Feasible Generalized Least square estimator in many
circumstances (Robert and Haichun 2011).
Long run Relationship: Cointegration Test

Cointegration one of the statistical property of time series variables
which assume that the stochastic process if integrated of order 1.
When a stochastic process is cointegrated, it suggest that there exist a
long run associativity or relation of the variable. Stochastic process of
Xt and yt are said to be cointegrated if there exists a α parameter such
that :µt = yt − αxt is a stationary process(Sorensen 2005)
Kao Residual Cointegration Test.

Null Hypothesis: No cointegration : Trend assumption: No deterministic trend

Newey-West automatic bandwidth selection and Barlett Kernel.
Test
t-Statistics
Prob
ADF
-4.117643
0.0000
Residual Variance
0.000607
HAC variance
0.000618
From above, the probability value of (0.0000) suggest the rejection of the null hypothesis
that the stochastic process is not cointegrated and implies that the stochastic process is
cointegrated. This imply also that exist a long run relationship of the series.
Recommendation

The generation of C02 emission is stated as the function of the population *GDP
per capital (GDP/Person).

The growth balanced path speed at 4.85% is very low compared to the desired
perfect convergence speed 1. Therefore, it suggest capital will leave South Africa
to Nigeria but at a slow speed. The reliance on the capital movement from South
Africa is anti-economics progressive. Countries with low initial capital like
Nigeria compare to South Africa should inject their capital and avoid capital
flight in their economy for a speedy convergence.
Recommendation Cont’

(Darmstadter 2001) states that the knowledge of the damages associated with
CO2 would serve as a useful clue in assessing the benefits of damages reduction.

Therefore implication of the negative externality from the industrial and
agriculture pose the economy with some ill effects like unhealthy atmospherics
space and climate change effects caused by the anthropogenic activities like
deforestation and fossil fuel combustion. Clean and renewable energy usage in
the sector should be advocated for and piguovian(1929) tax regime employed
with other environmental prohibition.
Recommendation Cont’

The positive relationships of Energy sector and waste
sector emission suggest that the economic activities in
the area adds to the productivity but not at its optimal
level since not significant. Investment in energy sector is
advised.
Conclusion.

While the global interest is on energy transition, the goal to drive productivity in
emerging economies must not suffer from the targets of emission since the effects
of greenhouse gases emission in Nigeria and South Africa productivity are not
significant and by extension in Africa.

A search for pareto point is inevitable for the negatively correlated greenhouse
gases emitting sectors with productivity.

Emerging economies should invest more in technology while countries with
lower initial capital should increase the inward capital investment.
Thank You for Listening