Table 1 Variables of the study
Symbol
|
Description of Variables
|
Unit of the variables
|
Source
|
CO2
|
Carbon emission
|
CO2 emissions (metric tons per capita)
|
WDI
|
EF
|
Ecological Footprints
|
Ecological consumption per capita (total)
|
GFP
|
MCAP
|
Market Capitalization
|
Capitalization of listed companies (% of GDP)
|
WDI
|
GDP
|
Gross Domestic Product
|
GDP (constant 2010 US$ per capita)
|
WDI
|
EC
|
Energy usage
|
Energy use (kg of oil equivalent per capita)
|
WDI
|
NREC
|
Non-Renewable Energy
|
Energy use (Exajoules)
|
BPS
|
URB
|
Urbanization
|
Population (Urban % of the total population)
|
WDI
|
EXP
|
Exports
|
Exports of products (% of GDP)
|
WDI
|
WDI, GFPN, & BPS stand for World Development Indicators (World Bank), Global Footprint Network, and BP Statistics.
Table 2 Descriptive statistics
Variables
|
N
|
M
|
SD
|
Min
|
Max
|
Skewness
|
Kurtosis
|
CO2
|
48
|
0.424
|
0.275
|
-0.026
|
0.878
|
0.296
|
1.949
|
EF
|
48
|
0.262
|
0.177
|
0.014
|
0.570
|
0.488
|
1.924
|
MCAP
|
48
|
2.015
|
0.325
|
1.245
|
2.470
|
-0.368
|
2.144
|
GDP
|
48
|
3.044
|
0.487
|
2.359
|
3.866
|
0.158
|
1.706
|
EC
|
48
|
2.964
|
0.223
|
2.656
|
3.363
|
0.576
|
2.031
|
RSS
|
48
|
1.555
|
0.350
|
0.929
|
2.101
|
0.110
|
1.852
|
URB
|
48
|
1.475
|
0.177
|
1.235
|
1.763
|
0.084
|
1.657
|
EXP
|
48
|
1.115
|
0.323
|
0.396
|
1.557
|
-0.642
|
2.257
|
N and M stand for the number of observations and mean of the variables used in this study. While CO2, EF, MCAP, EC, GDP, EXP, NREC, and URB stand for carbon emissions, ecological footprints, market capitalization, energy consumption, gross domestic product, exports, non-renewable energy consumption, and urbanization, respectively as reported in Table 1.
Table 3 Outcomes of unit root tests
Variables
|
ADF
|
|
ZA
|
I(0)
|
I(1)
|
|
I(0)
|
Y
|
N
|
I(1)
|
Y
|
N
|
CO2
|
-0.602
|
-4.2140***
|
|
-3.341
|
1990
|
30
|
-4.8960**
|
2010
|
41
|
EF
|
0.918
|
-4.3390***
|
|
-3.119
|
1998
|
29
|
-4.9040**
|
2010
|
41
|
MCAP
|
-1.738
|
-8.2390***
|
|
-4.3610*
|
2005
|
36
|
-6.8960***
|
2002
|
33
|
GDP
|
2.330
|
-3.9410***
|
|
-3.376
|
1980
|
11
|
-5.2220***
|
1986
|
16
|
EC
|
1.190
|
-3.6540***
|
|
-3.264
|
1998
|
29
|
-4.2070*
|
2010
|
41
|
RSS
|
-1.371
|
-4.0590***
|
|
-2.646
|
1978
|
9
|
-4.5360**
|
2006
|
37
|
URB
|
2.473
|
-3.0250**
|
|
-4.3370*
|
2009
|
40
|
-5.1150***
|
1980
|
11
|
EXP
|
-2.745
|
-5.2510***
|
|
-3.631
|
2006
|
37
|
-5.8820***
|
2004
|
35
|
URB
|
2.473
|
-3.0250**
|
|
-4.3370*
|
2009
|
40
|
-5.1150***
|
1980
|
11
|
All the variables are in logarithm form. ZA and ADF stand for Zivot & Andrews and Augmented Dickey-Fuller, respectively. Y and N is the year and observation number at which t-statistics is minimum while I (0) and I (1) indicate stationary at the level and 1st difference. *, **, *** demonstrate significance level at 10%; 5% and 1%, respectively.
Source: Author’s estimation.
Table 4 ARDL model- Lag selection
|
CO2
|
EF
|
MCAP
|
GDP
|
EC
|
EXP
|
NREC
|
URB
|
LR
|
8.9446*
|
7.1515*
|
73.7000*
|
8.5374*
|
13.1830*
|
143.5400*
|
22.5190*
|
6.2434*
|
FPE
|
0.0005*
|
0.0002*
|
0.0192*
|
0.0002*
|
0.0002*
|
0.0031*
|
0.0002*
|
2.4e-06*
|
AIC
|
-4.8105*
|
-5.9493*
|
-1.1133*
|
-5.9037*
|
-5.6351*
|
-2.9548*
|
-5.7966*
|
-10.1136*
|
SBIC
|
-4.6888*
|
-5.8277*
|
-1.0322*
|
-5.7820*
|
-5.5134*
|
-2.8737*
|
-5.6750*
|
-9.9109*
|
HQIC
|
-4.7654*
|
-5.9042*
|
-1.0832*
|
-5.8586*
|
-5.5899*
|
-2.9247*
|
-5.7515*
|
-10.0384*
|
p-value
|
0.0030
|
0.0070
|
0.0000
|
0.0030
|
0.0000
|
0.0000
|
0.0000
|
0.0120
|
Decision
|
I (1)
|
I (1)
|
I (1)
|
I (1)
|
I (1)
|
I (1)
|
I (1)
|
I (1)
|
All the variables are in logarithm form. FPE, LR, AIC, HQIC, and SBIC represent sequentially Final prediction error, modified LR test statistic, Akaike information criterion, Hannan-Quinn information criterion, and Schwarz Bayesian information criterion, respectively. Moreover, * shows the selected lag based on the p-value for each selection criteria.
Source: Author’s estimation.
Table 5. Bounds testing for ARDL
Test statistics
|
|
|
Value
|
k
|
F-statistics
|
|
|
20.9330
|
6
|
t-statistics
|
|
|
-1.8730
|
6
|
|
F-statistics
|
t-statistics
|
Significance level
|
Lower limit
|
Upper limit
|
Lower limit
|
Upper limit
|
10%
|
2.31
|
3.51
|
-2.57
|
-4.04
|
5%
|
2.73
|
4.06
|
-2.86
|
-4.38
|
1%
|
3.66
|
5.33
|
-3.43
|
-4.99
|
The lower limit and upper limit show the lower and upper boundary limits, respectively.
Source: Author’s estimation
Table 6 Dynamic ARDL Simulations (Carbon Emissions)
Variable Name
|
Coefficients
|
Standard Error
|
t-Statistics
|
p-value
|
Cons
|
-0.5488
|
0.2564
|
-2.1400
|
0.0400
|
MCAPt
|
-0.0449
|
0.0137
|
-3.2700
|
0.0020
|
∆MCAP t-1
|
0.0589
|
0.0125
|
4.7100
|
0.0000
|
GDPt
|
0.2593
|
0.1664
|
1.5600
|
0.1290
|
∆GDP t-1
|
-0.2473
|
0.1746
|
-1.4200
|
0.1660
|
ECt
|
0.5167
|
0.1831
|
2.8200
|
0.0080
|
∆EC t-1
|
-0.2882
|
0.1905
|
-1.5100
|
0.1400
|
NRECt
|
0.6374
|
0.1894
|
3.3700
|
0.0020
|
∆NREC t-1
|
-0.5884
|
0.1807
|
-3.2600
|
0.0030
|
URBt
|
-0.6072
|
0.5913
|
-1.0300
|
0.3120
|
∆URB t-1
|
0.4319
|
0.5965
|
0.7200
|
0.4740
|
EXPt
|
0.0612
|
0.0345
|
1.7800
|
0.0850
|
∆EXPt-1
|
-0.0032
|
0.0312
|
-0.1000
|
0.9180
|
ECT(-1)
|
-0.1991
|
0.1063
|
-1.8700
|
0.0700
|
R2
|
|
|
|
0.5834
|
Adj. R2
|
|
|
|
0.5375
|
N
|
|
|
|
47
|
Simulations
|
|
|
|
5000
|
Prob.
|
|
|
|
0.0000
|
F-Statistics
|
|
|
|
19.24
|
RMSE
|
|
|
|
0.0091
|
All the variables are in logarithm form. MCAP is market capitalization, GDP indicates gross domestic product, EC is energy consumption, NREC shows non-renewable energy consumption and URB is urbanization, and EXP displays exports. ∆ Represents difference while N stands for the number of observations used. ECT is an error correction term, RMSE is the root mean square error, while R2 shows the goodness of fit of the model.
Source: Author’s estimation.
Table 7 Dynamic ARDL Simulations (Ecological Footprints)
Variable Name
|
Coefficients
|
Standard Error
|
t-Statistics
|
p-value
|
Cons
|
-1.5530
|
0.4253
|
-3.6500
|
0.0010
|
MCAPt
|
-0.0176
|
0.0111
|
-1.5800
|
0.1240
|
∆MCAP t-1
|
-0.0105
|
0.0102
|
-1.0300
|
0.3110
|
GDPt
|
0.3788
|
0.1462
|
2.5900
|
0.0140
|
∆GDP t-1
|
-0.1300
|
0.1361
|
-0.9500
|
0.3470
|
ECt
|
0.7459
|
0.1476
|
5.0500
|
0.0000
|
∆ECt-1
|
-0.0785
|
0.1696
|
-0.4600
|
0.6470
|
NRECt
|
-0.2541
|
0.1688
|
-1.5100
|
0.1420
|
∆NRECt-1
|
0.0047
|
0.1358
|
0.0300
|
0.9720
|
URBt
|
0.0116
|
0.4496
|
0.0300
|
0.9800
|
∆URBt-1
|
-0.4148
|
0.4819
|
-0.8600
|
0.3960
|
EXPt
|
0.0084
|
0.0269
|
0.3100
|
0.7560
|
∆EXP t-1
|
0.0357
|
0.0267
|
1.3300
|
0.1910
|
ECT(-1)
|
-0.7567
|
0.1847
|
-4.1000
|
0.0000
|
R2
|
|
|
|
0.6659
|
Adj. R2
|
|
|
|
0.6537
|
N
|
|
|
|
47
|
Simulations
|
|
|
|
5000
|
Prob.
|
|
|
|
0.0000
|
F-Statistics
|
|
|
|
8.3100
|
RMSE
|
|
|
|
0.0071
|
All the variables are in logarithm form. MCAP is market capitalization, GDP indicates gross domestic product, EC is energy consumption, NREC shows non-renewable energy consumption and URB is urbanization, and EXP displays exports. ∆ Represents difference while N stands for the number of observations used. ECT is an error correction term, RMSE is the root mean square error, while R2 shows the goodness of fit of the model.
Sources: Author’s estimation.