The Asymmetric Impact of Informal Economy in the Energy-Economic
Growth Nexus in Saudi Arabia
ZOUHEYR GHERAIA
Department of Business Management, College of Business,
Jouf University, Skaka,
SAUDI ARABIA
HANANE ABDELLI
Department of Business Administration, College of Business,
Jouf University,
SAUDI ARABIA
RAJA HAJJI
Department of Quantitative methods, College of Business,
Sousse University,
TUNISIA
MEHDI ABID
Department of Business Management, College of Business,
Jouf University, Skaka,
SAUDI ARABIA
Abstract: At the macroeconomic level, the question of the informal sector is the most debated. This paper
studies the relationship between the informal economy (IFGDP), formal economy (FGDP), total economy
(TGDP), and energy consumption (EC) in Saudi Arabia. The Nonlinear Distributed Autoregressive Model
(NARDL) is used as an estimation technique on annual data ranging from 1970 to 2017. The empirical results
confirm the relationships between variables that are asymmetric. Positive and negative shocks on FGDP, TGDP
and IFGDP have positive effects on EC. The results will help policymakers and government officials have a
better understanding of the effect of the IFGDP on energy demand and FGDP in Saudi Arabia's development.
Keywords: Informal economy; Energy consumption; Saudi Arabia; Asymmetries.
Received: May 27, 2022. Revised: December 18, 2022. Accepted: January 19, 2023. Published: February 17, 2023.
1 Introduction and Background
In recent decades, energy consumption and
macroeconomic variables are examined in several
studies. The relationship between the informal
economy and energy consumption has received
little attention in theoretical as well as empirical
literature, [1], [2], [3], [4], [5]. One of the most
important findings is that the IFGDP accounts for a
large share of FGDP, especially for developing
countries6]. Without considering unrecorded
income when investigating the causal link between
energy consumption and economic growth, the
results may be biased. According to a recent article
by [7], 157 countries between 1991 and 2017 were
analyzedto determine the size and growth of the
IFGDP.For the total sample, the informal economy
accounts for 30.9% of GDP. It is estimated that the
informal economy in Saudi Arabia makes up about
17% of its formal economy. It is suggested that
two-thirds of IFGDP would be spent on the FGDP,
[8], [9]. Empirical and theoretical studies indicate
that the underground economy reduces real GDP
because of the lack of tax revenue. Several studies
showed that small informal firms, such as [10] are
unproductive, rarely become formal, and pay less
than half as much as small formal firms.
Empirical studies examine the relationship between
EC, FGDP and the IFGDP are very rare, compared
to a vast literature that studies the relationship
between energy and growth. A pioneering study by
[1] estimates the informal economy in Turkey
between 1973 and 2003 using a methodology
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.41
Zouheyr Gheraia, Hanane Abdelli,
Raja Hajji, Mehdi Abid
E-ISSN: 2224-2899
454
Volume 20, 2023
developed. They are pioneers in the idea that
informal economies can be measured by CO2
emissions and energy consumption. Taking into
account the size of the informal economy, [2]
examines the long-term impact of EC on TFGDP in
Turkey during the period 1970-2005. They show
that the relationship between TGDP and EC is
rejected in the long-term, while EC strongly
influences the FGDP. In contrast, FGDP and EC
are found to be causally related in the short-term,
but TGDP and EC do not appear to have any causal
relationship.EC-FGDP nexus therefore supports the
conservative hypothesis, implying that energy
conservation policies can reduce greenhouse gas
emissions without affecting economic growth.
IFGDP and production are found to be unstable
over time. The later proves the presence of the
neutrality hypothesis. Consequently, the
implementation of economic policies aimed at
reducing the IFGDP cannot serve as a complement
to energy conservation programs.
From 1980 to 2009, [11] examines the causal
relationship between FGDP and EC in Tunisia in
the presence of the IFGDP. The empirical results
indicate that there is Granger causality running
from EC to FGDP and TGDP. To reduce the
number of polluting emissions, the government
must use more effective instruments. This analysis
suggests that informal economic growth contributes
significantly to environmental degradation, which
has important policy implications. Between the
years 1980-2012, [12] studies the relationship for
159 countries between IFGDP and EC. Their
results are reported for several groups of countries
based on their informal economies. The IFGDP
negatively impacts EC, according to their findings.
In emerging countries, for example, the size of the
informal sector increased by 1%, resulting in a
decrease in energy intensity of about 0.13%.
Furthermore, the relationship between IFGDP and
EC is U-shaped. In particular, all countries whose
IFGDP is less than 20% of their FGDP showed a
negative relationship between EC and the IFGDP.
The impact of IFGDP on environmental pollution
in African countries from 1991 to 2015 is examined
by [4]. They found that the IFGDP and institutional
quality are significant contributors to
environmental pollution in Africa by using ordinary
least squares, fixed effects, and generalized system
method of moments. Furthermore, the IFGDP
influences institutional quality in the region, which
in turn deteriorates the quality of the environment.
According to this information, the low level of
institutional quality in the region leads to a higher
level of IFGDP, and therefore a greater degree of
environmental pollution. Recently, [5] examines
the relationship between the IFGDP and the
ecological footprint for the case of Africa during
the 1991-2017 periods. The study finds that both
the IFGDP and FGDP have positive and
statistically significant impacts on ecological
footprints, suggesting that the IFGDP and FGDP
contribute to environmental degradation. In similar
studies, [13] analyzes data from South Asian
countries to study the effect of IFGDP on EC and
pollution. The study shows increased EC in Sri
Lanka and Pakistan, but decreased EC in India
when using the Autoregressive Distributed Lag
Model (ARDL). Thus, the Nonlinear ARDL
(NARDL) model shows that the IFGDP contributes
to the improvement of EC in Pakistan.
Furthermore, we add to the empirical literature in a
variety of fields in this context. Despite studies in
Saudi Arabia ignoring the role of the IFGDP on
EC, this study is the first to examine its effect on
EC, [14], [15], [16]. Second, this study is needed
since the impact of the IFGDP on EC is neglected.
FGDP cannot be used alone to understand the
affect of economic activities on EC.
The reason for using Saudi Arabia in this study is
as follows: In the Middle East, Saudi Arabia has
the largest economy and is the richest Arab nation.
By implementing a major public works policy,
attracting foreign direct investment, and ensuring a
sound banking and financial system, the country
has become the largest economy in the region.
However, Saudi Arabia suffers from a phenomenon
that threatens its economy, which is informal
economy. This type of economy constitutes an
important part of the GDP volume, since the rate of
informal economy in the Kingdom during the
period 1991-2017isestimated at 16.28% of the
volume of GDP, [6]. Such economic growth is
almost entirely based on oil and gas, which has an
impact on the country's environmental
sustainability.
Following is an outline of the remainder of the
paper. The data and methodology are presented in
section 2. Section 3 presents the empirical results,
and the conclusion and policy recommendations are
discussed in the final section.
2 Data and Methodology
2.1 Data
The data includes four variables of the Saudi
economy, namely energy consumption (EC),
formal gross domestic product (FGDP), total gross
domestic product (TGDP) and informal economy
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.41
Zouheyr Gheraia, Hanane Abdelli,
Raja Hajji, Mehdi Abid
E-ISSN: 2224-2899
455
Volume 20, 2023
(IFGDP), and covers the period 1970-2017. Data
for FGDP and EC are taken from the World Bank
database (WDI, 2022). The IFGDP data are taken
from the articles of [6] and [17]. The FGDP is
expressed in dollars (US constant 2015). The
TGDP is the sum of FGDP and IFGDP. EC is
expressed in kilograms of oil equivalent per capita.
Figure 1 describes the trajectory of our variables.
The graph shows that all variables follow an
upward trend and increase during the examined
period. FGDP and TGDP show a common trend
over the entire period. The difference between
TGDP, FGDP and IFGDP seems to have a similar
shape as FGDP and TGDP in the time period
considered.
7.6
8.0
8.4
8.8
9.2
9.6
10.0
10.4
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
LnEC lnIFGDP LnFGDP lnTGDP
Fig. 1: Energy Consumption, FGDP, TGDP and IFGDP.
Table 1 presents the descriptive statistics and the
stochastic properties of the variables used in our
study. Based on our results, we show that TGDP
and EC have dispersion coefficients of 0.06 and
0.218, respectively. With the exception of energy
consumption, all the series have positively
asymmetric distributions, which means their lines
are longer than those in a normal distribution. The
IFGDP and the TGDP show excess kurtosis,
indicating that they have fatter tails than a normal
distribution. Data variables are normally distributed
according to the Jarque-Bera statistic.
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.41
Zouheyr Gheraia, Hanane Abdelli,
Raja Hajji, Mehdi Abid
E-ISSN: 2224-2899
456
Volume 20, 2023
Table 1. Summary statistics for the series.
IFGDP
FGDP
TGDP
Mean
8.001219
9.814599
9.966765
Median
8.019457
9.816604
9.971686
Maximum
8.186980
9.934401
10.07937
Minimum
7.810439
9.624406
9.793305
Std. Dev.
0.105242
0.079093
0.068511
Skewness
-0.261067
-0.471811
-0.585000
Kurtosis
2.171914
2.715220
3.039396
Jarque-Bera
1.277802
1.295362
1.827270
Probability
0.527872
0.523258
0.401064
Sum
256.0390
314.0672
318.9365
Sum Sq. Dev.
0.343350
0.193926
0.145508
Observations
48
48
48
2.2 Methodology
According to previous studies, the ARDL
methodology proves insufficient to analyze both
long-term and short-term relationships between
variables when the dynamics of those variables
show nonlinear patterns, [18], [3], [19], [20], [21].
Non-linearity is commonly observed in economic
and financial time series for a variety of reasons.
Indeed, economic and financial time series are less
likely to follow simple linear paths during the
period when we conducted our research, because
several events complicate them. In fact, The Asian
economic crisis of 1997, the oil shocks in 2008, and
the global financial crisis of 2008 were among the
most important events during the period 1980-
2017. It is necessary to develop even more
sophisticated models in order to obtain robust
results after sudden events cause structural breaks
in time series data. In our study, we used [22]
nonlinear ARDL model. This model can
incorporate long- and short-term asymmetries as
well as non-linearity, while simultaneously taking
into account the cointegration between variables in
the model. Formally, the linear ARDL model has
the following form:
11
10
11
00
1 1 1
1
pq
t j t j j t j
jj
qq
j t j j t j
jj
EC t FGDP t IFGDP t
TGDP t t
EC EC FGDP
IFGDP TGDP
EC FGDP IFGDP
TGDP










(1)
According to Akaike and Schwarz's information
criteria, p and q represent delay orders. The symbol
Δ represents the first difference operator.
Based on the simultaneous study of the long- and
short-term asymmetry effects of the ARDL model
above along with the evaluation of NARDL models
for each variable, the following three NARDL
models are estimated:
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.41
Zouheyr Gheraia, Hanane Abdelli,
Raja Hajji, Mehdi Abid
E-ISSN: 2224-2899
457
Volume 20, 2023
11
10
1
()
pq
t j t j i t j i t j
jj
EC t FGDP t j FGDP t j t
EC EC FGDP FGDP
EC FGDP FGDP



(2)
11
10
1
()
pq
t j t j i t j i t j
jj
EC t TGDP t j TGDP t j t
EC EC TGDP TGDP
EC TGDP TGDP



(3)
11
10
1
()
pq
t j t j i t j i t j
jj
EC t IFGDP t j IFGDP t j t
EC EC IFGDP IFGDP
EC IFGDP IFGDP



(4)
Positive and negative partial sums are denoted by
(+) and (-) in Eqs (2)-(4) and are calculated as
follows:
11
FGDP = max(0, )
tt
t j j
jj
FGDP FGDP



and
11
FGDP = min( ,0);
tt
t j j
jj
FGDP FGDP



11
TGDP = max(0, )
tt
t j j
jj
TGDP TGDP



and
11
TGDP = min( ,0);
tt
t j j
jj
TGDP TGDP



11
IFGDP = max(0, )
tt
t j j
jj
IFGDP IFGDP



and
11
IFGDP = min( ,0);
tt
t j j
jj
IFGDP IFGDP



For every determinant of energy consumption,
positive and negative coefficients are calculated
similarly. For example, the long-term positive and
negative coefficients for FGDP are calculated as
FGDP
FGDP
EC

and
FGDP
FGDP
EC

,respectively. Using Wald statistics, we test the
long- and short-term asymmetry of the NARDL
models in equations (2)-(3). We use a Wald statistic
test for long-term asymmetry in energy
consumption for each determinant Y (FGDP,
TGDP, and IFGDP) with the null hypothesis:
YY


.For short-term symmetry, we use a Wald
statistic for null hypotheses as follows:
ii


for
1,2,..., 1.iq
If the Wald test allows the null
hypothesis of long- or short-term symmetry to be
accepted for a determinant of energy consumption,
linearity is imposed for that particular variable and
the associated constrained NARDL model is
estimated.
In the event that asymmetries are detected (long-
term or short-term), the following formulas are
used to calculate the asymmetrical multipliers for
each determinant Y(FGDP, TGDP and IFGDP) on
changes in EC (positive or negative variations):
,0
htj
hY jt
EC
mY
and
,0
htj
hY jt
EC
mY
Shin et al., [22] showed that
,h Y Y
m

and
,h Y Y
m

, knowing that
h
.
3 Empirical Results
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.41
Zouheyr Gheraia, Hanane Abdelli,
Raja Hajji, Mehdi Abid
E-ISSN: 2224-2899
458
Volume 20, 2023
It is necessary to test for (non) stationarity by using
both the ADFand ZA unit root tests, which are
more appropriate for nonlinear series if breaks are
present in their trajectory. ZA and ADF unit root
tests can be found in Table 2 below.
Table 2. Unit Roots tests
Series
ADF
ZA
Levels
First difference
Levels
First difference
EC
-5.145***
_
-10.655***
_
FGDP
0.235
-6.587***
-2.364
-7.633***
IFGDP
0.128
-2.364**
-3.928
-8.099***
TGDP
0.556
-2.927**
-3.310
-5.365**
Note: The critical values of the ZA(1992) test for 1%, 5%, and 10% significance levels are 5.57,
5.08, and 4.82.
Table 2 shows that all the variables, with the
exception of EC, cannot be rejected by the null
hypothesis of non-stationarity. These tests reject
non-stationarity for all variables of first-difference,
indicating that all variables are I(1), except EC.
There is a difference between the order of
integration of EC compared to other variables since
energy consumption is stationary in level (I(0)).As
a result of variable stationarity, Johansen's
cointegration method cannot be used to test
whether the variables have a common long-term
relationship since it requires that the variables are
integrated equally. It appears that the Johansen
cointegration test is not appropriate when there is a
difference in the order of integration between the
variables. Therefore, in order to test if the variables
are cointegrated, we use the NARDL methodology.
Using Wald statistics, the results of the tests for
long- and short-term asymmetries are presented in
Table 3.There is no evidence that long- and short-
term asymmetries exist in FGDP, TGDP, and
IFGDP according to Wald statistics. On the long
and short term, these results demonstrate a
nonlinear and asymmetrical response of energy
consumption to FGDP, TGDP, and IFGDP. The
long-term relationship between the underlying
variablesisconfirmed by the presence of short-and
long-term asymmetries. In order to accomplish this,
we apply the [22] nonlinear test approach.
Table 3. Wald Test for Short- and Long-term Symmetries
Wald test
FGDP
TGDP
IFGDP
WLR
8.238*
12.781**
3.088***
WSR
10.022*
15.134*
6.158*
Note: The Wald test for short-term symmetry is represented by the WSR.The Wald test for long-
term symmetry is represented by the WLR. *, **, and *** indicate rejecting the null hypotheses of
short- and long-term symmetry at the levels of significance of 1%, 5%, and 10%, respectively.
According to Table 4, the bounds tests for
asymmetric cointegration produced the following
results. We use the T(TBDM) statistic developed
by [23] as well as the F statistic (SPSS) developed
by [24] to investigate whether there is nonlinearity.
The calculated F-statistics of [24] and the BDM test
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.41
Zouheyr Gheraia, Hanane Abdelli,
Raja Hajji, Mehdi Abid
E-ISSN: 2224-2899
459
Volume 20, 2023
t-statistics are greater than the upper critical value,
rejecting the null hypothesis that there is no
asymmetric cointegration. This is more evident in
the FGDP, the TGDP, and the IFGDP. For the
Saudi economy, it may be better to introduce a
measure of IFGDP in the analysis of EC and
economic growth over the long and short term.
According to the empirical findings, the EC,
FGDP, TGDP, and IFGDP have long-term
asymmetric relationships.
Table 4. Bound Testing for Asymmetric Cointegration
FGDP
TGDP
IFGDP
FPSS
25.366*
30.562*
27.304*
TBDM
-4.358*
-6.254*
-3.774***
2
Normal
3.012
1.981
4.011
2
ARCH
0.224
0.337
0.207
2
RESET
0.501
0.286
0.422
2
SERIAL
1.336
0.885
0.905
Pesaran et al., (2001)
Banerjee et al., (1998)
Significance level
LCB I(0)
UCB I(1)
Significance level
Critical values
1%
3.271
5.365
1%
-4.713
5%
2.636
3.551
5%
-4.035
10%
2.331
3.224
10%
-3.678
Note: * and *** denote significance at the 1% and 10% levels, respectively.
Following the results in Table 3, which confirm the
existence of the short- and long-term asymmetric
relationships between EC and its determinants in
Saudi Arabia, we estimate the NARDL models
given in equations (2)-(3) to verify the asymmetric
effect of FGDP, TGDP and IFGDP on long-term as
well as short-term energy consumption. Using the
three models above, we can estimate the NARDL
in Table 5. As a potential determinant of EC, we
consider FGDP in the first model. In the next
models, we replace the FGDP by the TGDP, and
then we investigate its impact on the EC using the
IFGDP. Based on the empirical results, all three
models considered have negative lagged dependent
variables that are statistically significant, which
confirms the model’s stability condition.
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.41
Zouheyr Gheraia, Hanane Abdelli,
Raja Hajji, Mehdi Abid
E-ISSN: 2224-2899
460
Volume 20, 2023
Table 5. NARDL estimation results
FGDP
TGDP
IFGDP
Variables
Coefficient
t-Statistic
Variables
Coefficient
t-Statistic
Variables
Coefficient
t-Statistic
Constant
1.351***
7.022
Constant
1.982***
8.011
Constant
2.367***
5.667
1t
FGDP
0.258***
5.697
1t
TGDP
0.492***
6.354
1t
IFGDP
0.265***
6.384
1t
FGDP
0.189*
1.801
1t
TGDP
0.365***
6.064
1t
IFGDP
0.198**
2.088
t
FGDP
_
_
t
TGDP
_
_
t
IFGDP
0.169*
1.762
1t
FGDP
-0.656***
6.022
1t
TGDP
-0.305***
5.055
1t
IFGDP
-0.681***
7.881
2t
FGDP
-0.662***
5.984
2t
TGDP
-0.368***
4.964
2t
IFGDP
-0.627**
2.681
t
FGDP
-0.684***
4.224
t
TGDP
-0.299**
2.337
t
IFGDP
_
_
1t
FGDP
-0.506**
2.354
1t
TGDP
-0.247***
5.972
1t
IFGDP
_
_
Note: A positive partial sum is represented by a superscript "+", whereas a negative partial sum is represented by a
superscript "-". ***, **, and *denote significance at the 1%, 5% and 10% levels.
EC is positively affected by positive (negative)
shocks to the FGDP, TGDP, and IFGDP, in the
long term. EC is more influenced by these
variables’ increases than their decreases. Using the
NARDL, the independent variables are
decomposed into positive and negative partial
sums. Increased EC results from positive changes
in FGDP, TGDP, and IFGDP, in the long term. On
the other hand, a decrease in FGDP, TGDP, and
IFGDP will result in a decrease in EC, in the long
term.
In Table 5, a 1% change in FGDP leads to an
increase in EC by 0.258% for the dependent
variable EC. In contrast, the EC increases by
0.189% when FGDP's partial function changes
negatively. With a change of 1%, the positive
changes in the cumulative function of FGDP and
the negative changes in the partial function of
FGDP decrease EC by 0.645%, in the short term.
Also, the cumulative function also increases EC by
0.492% for a 1% change in long-term TGDP.
Additionally, for a 1% change in TGDP, EC
increases by 0.365% if there is a negative change in
the cumulative function. In the short term,
however, for a 1% change in TGDP, negative and
positive changes in the cumulative function of
TGDP reduce EC by 0.305% and 0.299%,
respectively. It is predicted that the cumulative
function of IFGDP increases EC by 0.26 % for a
1% change in IFGDP in the long-term. In contrast,
a decrease in IFGDP's partial function increases EC
by 0.198%. A 1% change in IFGDP, however,
decreases EC by 0.681% and 0.627% for positive
and negative changes in IFGDP’s cumulative
functions, respectively.
The validation of our estimated model as well as
the results obtained from the short- and long-term
relationship requires the verification of a set of
hypotheses, namely error correlation,
heteroscedasticity, normality, specification and
coefficient stability. Indeed, the four tests presented
in Table 5 show that the probability of the statistic
for each test is greater than 5%. This means that the
null hypothesis is accepted in all these tests. The
errors are therefore not autocorrelated,
homoscedastic, their distribution follows a normal
law and our model is well specified. In addition, the
stability of the coefficients of our ARDL model is
validated through the CUSUM and CUSUMSQ
tests, since the curve does not go out of the corridor
in these two tests (Figures 2-4). Finally, based on
the results of the five tests performed, we can
confirm the robustness of our estimated NARDL
model.
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.41
Zouheyr Gheraia, Hanane Abdelli,
Raja Hajji, Mehdi Abid
E-ISSN: 2224-2899
461
Volume 20, 2023
-15
-10
-5
0
5
10
15
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
CUSUM 5% Significance
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
CUSUM of Squares 5% Significance
Fig. 2: CUSUM and CUSUM of Square for Formal GDP.
-15
-10
-5
0
5
10
15
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
CUSUM 5% Significance
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
CUSUM of Squares 5% Significance
Fig. 3: CUSUM and CUSUM of Square for Informal GDP.
-15
-10
-5
0
5
10
15
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
CUSUM 5% Significance
-0.4
0.0
0.4
0.8
1.2
1.6
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
CUSUM of Squares 5% Significance
Fig. 4: CUSUM and CUSUM of Square for Total GDP.
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.41
Zouheyr Gheraia, Hanane Abdelli,
Raja Hajji, Mehdi Abid
E-ISSN: 2224-2899
462
Volume 20, 2023
Table 5. Robustness test
Statistics
FGDP
TGDP
IFGDP
Breusch-Godfrey
0.467
0.525
0.398
ARCH
0.189
0.357
0.584
Jarque-Bera
0.654
0.544
0.365
Ramsey
0.288
0.354
0.521
CUSUM
Stable
Stable
Stable
CUSUMsq
Stable
Stable
Stable
The NARDL model also presents the long-term
asymmetric response of EC in Saudi Arabia to
positive and negative variations of its determinants,
respectively, after analyzing the long- and short-
term impacts of FGDP, TGDP, and IFGDP on EC.
Table 6 presents the long-term skew parameters for
the three estimated models. According to Table 4,
Saudi Arabia's FGDP, TGDP, and IFGDP affect
EC in the long term. Specifically, a 1% decrease in
FGDP, TGDP, and IFGDP raise energy
consumption by 2.35%, 2.52%, and 1.60% in our
estimated models, respectively. In other words,
official GDP, real GDP and the EU increase energy
consumption by 2.35%, 2.52% and 1.60% when
they increase by 1%. In contrast, our estimated
models estimate EC to be reduced by 0.36%,
0.43%, and 1.01% for a 1% decrease in FGDP,
TGDP, and IFGDP, respectively.
Table 6. Long-term parameters.
FGDP
TGDP
IFGDP
Coefficients
Statistics
Coefficients
Statistics
Coefficients
Statistics
FGDP
2.354***
TGDP
2.521***
IFGDP
1.609***
FGDP
0.365**
TGDP
0.434**
IFGDP
1.011**
FGDP
,
FGDP
,
TGDP
,
TGDP
,
IFGDP
and
IFGDP
represent estimated long-term asymmetric coefficients associated with
the change in FGDP, TGDP, and IFGDP, respectively. ***, **, *denote significance at the 1%, 5% and 10% levels.
Following a positive and negative unitary shock
destabilizing the economy, figures 5 to 7 illustrate
the trajectory of asymmetric adjustments to a new
long-term equilibrium. Up to an 80-period horizon,
the green and red dotted lines show how energy
consumption responds to a positive and negative
unitary shock. Positive and negative unit shocks are
represented by a blue curve that represents the
asymmetry line. In the blue area, we can see the
95% confidence interval for the asymmetry curve.
Using these figures, it is possible to predict how EC
will respond to an exogenous shock, either positive
or negative.
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.41
Zouheyr Gheraia, Hanane Abdelli,
Raja Hajji, Mehdi Abid
E-ISSN: 2224-2899
463
Volume 20, 2023
-1
0
1
2
3
4
5
1 3 5 7 9 11 13 15
Multiplier for LNFGDP(+)
Multiplier for LNFGDP(-)
Asymmetry Plot (with C.I.)
Fig. 5: Multipliers for Formal GDP Model.
-1
0
1
2
3
4
5
1 3 5 7 9 11 13 15
Multiplier for LNFGDP(+)
Multiplier for LNFGDP(-)
Asymmetry Plot (with C.I.)
Fig. 6: Multipliers for Informal GDP Model.
-1
0
1
2
3
4
5
1 3 5 7 9 11 13 15
Multiplier for LNFGDP(+)
Multiplier for LNFGDP(-)
Asymmetry Plot (with C.I.)
Fig. 7: Multipliers for Total GDP Model.
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.41
Zouheyr Gheraia, Hanane Abdelli,
Raja Hajji, Mehdi Abid
E-ISSN: 2224-2899
464
Volume 20, 2023
Indeed, for authorities and decision-makers, it is
crucial to have an accurate forecast of future EC. In
this way, the Saudi authorities can take the
necessary precautions to prevent a disruption of
energy supply that would both deteriorate
household life quality and disrupt the production
process. Alternatively, policymakers could control
the EC by monitoring its determinants and adopting
the necessary policies to limit its negative effects
on the environment, such as CO2 emissions and air
pollution, associated with EC. Figure 5 shows the
cumulative multipliers for EC and FGDP. It is
evident from the graph that EC is positively
associated with FGDP, and that negative shocks
dominate positive shocks in FGDP.A comparison
of figures 6 and 7 shows that the dynamic multiples
trail similar trajectories regardless of the economic
variables introduced into the model among TGDP
and IFGDP. For the first two years, positive unit
shocks were more effective than negative unit
shocks, and then negative shocks were greater than
positive unit shocks in affecting EC. In response to
unitary positive and negative shocks in TGDP and
IFGDP, the asymmetry curve follows a similar
pattern, starting with a significant negative reaction
in EC. The negative feedback reaches its peak after
about three quarters, and the new equilibrium path
for EC follows about six quarters.
4 Conclusion
In this article, we study the relationship between
the IFGDP and EC in Saudi Arabia. The study uses
annual frequency data for the period 1970-2017.
Furthermore, FGDP, IFGDP, and TGDP were
examined in relation to EC in Saudi Arabia. This
study investigates the long- and short-term effects
of FGDP, IFGDP, and TGDP on EC in Saudi
Arabia using the nonlinear autoregressive
distributed lag model developed by [22].
Based on the empirical results, the variables studied
exhibit an asymmetric cointegration relationship.
Specifically, the results reveal that in the long term,
positive changes in FGDP, TGDP and the IFGDP
lead to higher EC. Moreover, negative changes in
the FGDP, TGDP and IFGDP reduce EC in Saudi
Arabia in the long term. Based on the short-term
analysis, the increase in FGDP, TGDP, and IFGDP
reduces the short-term EC in Saudi Arabia. We
need to adopt new strategies that contribute to the
action plan while also respecting sustainable
development goals. It is pertinent to Saudi Arabia's
OPEC energy policy to address this challenge.
From a policy perspective, this result further
suggests that Saudi Arabian policymakers can
adopt effective policies to control long- and short-
term energy consumption through the informal
economy channel, in the sense that the fight against
the informal economy is not a priority for the Saudi
economy. Instead, in 2016, Mohammed Bin
Salman announced Vision 2030. The main
objective of the Vision 2030 plan is to ensure the
Kingdom's transition to a new model of economic
development, more liberal and more open to the
world, creating jobs and wealth. In terms of energy,
the Kingdom must substantially reduce its domestic
consumption. Arabia is, in fact, one of the largest
world consumers of black gold for domestic
purposes (nearly 3 billion barrels, or +6% per year
since 1940). With this in mind, the transition to
renewable and clean energies is a priority. The
Saudi energy target for 2030 is based on the
production of more than 58.7 GW of renewable
energy mainly combining solar and wind power. To
meet this challenge - and also to meet the expected
increase of more than 300% in electricity
consumption by 2030 - the city of Riyadh has
undertaken to make this sector more attractive and
more open to foreign investors, particularly with
privatizations and abolishing monopolies. This is,
in our opinion, a better solution because Saudi
Arabia has good hydroelectric potential, and this
resource is not yet fully exploited. Another effort in
Saudi Arabia seems to be made in the hydrocarbon
industry. Foreign and domestic investment in this
sector has enabled Saudi Arabia to more than
double its production of natural gas, one of the
main energy resources used by households and
businesses. Indeed, the production of natural gas is
a good channel to increase economic growth as
these exports accounted for about 12% of the
country's GDP in 2016.
Future research can still improve upon the findings
despite the significant methodological and policy
contributions. Itissuccessfully demonstrated that
energy consumption/formal GDP, energy
consumption/total GDP, and energy
consumption/informal GDP are non-linear
relationships. In the next research, we will analyze
the specific turning points of all the non-linear
relationships.
Acknowledgements:
This work is funded by the Deanship of Scientific
Research at Jouf University under Grant Number
(DSR2022-RG-0161)
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.41
Zouheyr Gheraia, Hanane Abdelli,
Raja Hajji, Mehdi Abid
E-ISSN: 2224-2899
465
Volume 20, 2023
References:
[1] Karanfil, F., & Ozkaya, A. (2007). Estimation of
real GDP and unrecorded economy in Turkey
based on environmental data. Energy
policy, 35(10), 4902-4908.
[2] Karanfil, F. (2008). Energy consumption and
economic growth revisited: Does the size of
unrecorded economy matter?. Energy
policy, 36(8), 3029-3035.
[3] Benkraiem, R., Lahiani, A., Miloudi, A., &
Shahbaz, M. (2019). The asymmetric role of
shadow economy in the energy-growth nexus in
Bolivia. Energy policy, 125, 405-417.
[4] Dada, J. T., Ajide, F. M., & Sharimakin, A.
(2021). Shadow economy, institutions and
environmental pollution: insights from
Africa. World Journal of Science, Technology
and Sustainable Development.
[5] Dada, J. T., Ajide, F. M., & Adeiza, A. (2022).
Shadow economy and environmental pollution in
West African countries: The role of
institutions. Global Journal of Emerging Market
Economies, 14(3), 366-389.
[6] Medina, L., & Schneider, F. (2019). Shedding
light on the shadow economy: A global database
and the interaction with the official one. Available
at SSRN 3502028.
[7] Hassan, M., & Schneider, F. (2016). Size and
development of the shadow economies of 157
countries worldwide: Updated and new measures
from 1999 to 2013. Available at SSRN 2861026.
[8] Schneider, F., & Enste, D. H. (2000). Shadow
economies: Size, causes, and
consequences. Journal of economic
literature, 38(1), 77-114.
[9] Williams, C. C., & Schneider, F.
(2016). Measuring the Global Shadow Economy:
the prevalence of informal work and labour.
Edward Elgar Publishing.
[10] La Porta, R., & Shleifer, A. (2014). Informality and
development. Journal of economic
perspectives, 28(3), 109-26.
[11] Abid, M. (2016). Energy consumption-informal
economic growth analysis: what policy options do
we have?. Journal of the Knowledge
Economy, 7(1), 207-218.
[12] Basbay, M. M., Elgin, C., & Torul, O. (2016).
Energy consumption and the size of the informal
economy. Economics, 10(1).
[13] Sohal, A., Nand, A. A., Goyal, P., & Bhattacharya,
A. (2022). Developing a circular economy: An
examination of SME’s role in India. Journal of
Business Research, 142, 435–447.
[14] Adedoyin, F. F., Ozturk, I., Agboola, M. O.,
Agboola, P. O., & Bekun, F. V. (2021). The
implications of renewable and non-renewable
energy generating in Sub-Saharan Africa: The
role of economic policy uncertainties. Energy
Policy, 150, 112115.
[15] Murshed, M., Apergis, N., Alam, M. S., Khan, U.,
& Mahmud, S. (2022). The impacts of renewable
energy, financial inclusivity, globalization,
economic growth, and urbanization on carbon
productivity: Evidence from net moderation and
mediation effects of energy efficiency
gains. Renewable Energy, 196, 824-838.
[16] Mezghani, I., & Haddad, H. B. (2017). Energy
consumption and economic growth: An empirical
study of the electricity consumption in Saudi
Arabia. Renewable and Sustainable Energy
Reviews, 75, 145-156.
[17] Elgin, C., & Oztunali, O. (2012). Shadow
economies around the world: model based
estimates. Bogazici University Department of
Economics Working Papers, 5(2012), 1-48.
[18] Abbasi, K., Jiao, Z., Shahbaz, M., & Khan, A.
(2020). Asymmetric impact of renewable and
non-renewable energy on economic growth in
Pakistan: New evidence from a nonlinear
analysis. Energy Exploration &
Exploitation, 38(5), 1946-1967.
[20] Namahoro, J. P., Wu, Q., Xiao, H., & Zhou, N.
(2021). The impact of renewable energy,
economic and population growth on CO2
emissions in the East African region: evidence
from common correlated effect means group and
asymmetric analysis. Energies, 14(2), 312.
[21] Mighri, Z., & AlSaggaf, M. I. (2022).
Asymmetric impacts of renewable energy
consumption and economic complexity on
economic growth in Saudi Arabia: evidence from
the NARDL model. Environmental Science and
Pollution Research, 1-28.
[22] Shin, Y., Yu, B., & Greenwood-Nimmo, M.
(2014). Modelling asymmetric cointegration and
dynamic multipliers in a nonlinear ARDL
framework. In Festschrift in honor of Peter
Schmidt (pp. 281-314). Springer, New York, NY.
[23] Banerjee, A., Dolado, J., & Mestre, R. (1998).
Error‐correction mechanism tests for
cointegration in a single‐equation
framework. Journal of time series analysis, 19(3),
267-283.
[24] Pesaran, M. H., Shin, Y., & Smith, R. J. (2001).
Bounds testing approaches to the analysis of level
relationships. Journal of applied
econometrics, 16(3), 289-326.
WSEAS TRANSACTIONS on BUSINESS and ECONOMICS
DOI: 10.37394/23207.2023.20.41
Zouheyr Gheraia, Hanane Abdelli,
Raja Hajji, Mehdi Abid
E-ISSN: 2224-2899
466
Volume 20, 2023
This work is funded by the Deanship of Scientific
Research at Jouf University under Grant Number
(DSR2022-RG-0161)
Contribution of Individual Authors to the
Creation of a Scientific Article (Ghostwriting
Policy)
The authors equally contributed in the present
research, at all stages from the formulation of the
problem to the final findings and solution.
Sources of Funding for Research Presented in a
Scientific Article or Scientific Article Itself
Conflict of Interest
The authors have no conflicts of interest to declare
that are relevant to the content of this article.
Creative Commons Attribution License 4.0
(Attribution 4.0 International, CC BY 4.0)
This article is published under the terms of the
Creative Commons Attribution License 4.0
https://creativecommons.org/licenses/by/4.0/deed.en
_US