Beruflich Dokumente
Kultur Dokumente
A R T I C LE I N FO A B S T R A C T
Keywords: The paper examines the electricity demand behavior for Saudi Arabia, using annual data for the period of
Electricity demand 1990–2016, at regional level. The study finds that income, price and population are the main drivers of elec-
Price elasticity tricity demand at regional level. Although, the impacts vary across regions (central, eastern, southern and
Cointegration western), the estimated elasticities all are statistically significant, in both long and short run, and have the
Regions
expected signs for all the regions. The income, price and population elasticities range from 0.10 to 0.93, from
Saudi Arabia
-0.63 to -0.06, and from 0.24 to 0.95, respectively, across regions in the long run. In the short run these intervals
are (0.05, 0.47), (−0.27, −0.01) and (0.13, 1.49) respectively for income, price and population across the
regions. The findings reported in this paper, should assist policy makers to develop insights about the potential
regional impact of changes to electricity prices, income and population patterns.
1. Introduction the electricity consumption shape it over time in the Central, Eastern,
Southern and Western regions of Saudi Arabia and suggest policy in-
Since energy is a pervasive input to all business and recreational sights accordingly.
activities, total energy demand is an important indicator that helps Note that this 4-region classification is adopted by the Saudi
explain the pattern of economic development within a country. Electricity Company (SEC) and sum of these four regions’ consumption
Identifying and understanding the key determinants of electricity de- add up to the total electricity consumption of the Kingdom.
mand is therefore important for the economic prosperity of a country, The following points motivated us to conduct electricity consump-
since the availability of reliable electricity directly affects the prospects tion analysis at the regional level. First, the regional dimension is im-
of sustainable economic development. With important mega projects portant because of the differences in weather patterns (and subse-
already in the works and national transformation programs such as the quently the electricity demand profiles) across the country. Second, the
National Industrial Development and Logistics Program (NIDLP, 2019) distribution of residential, commercial, and industrial activities is dif-
already being implemented, understanding the existing and projected ferent across the regions and this may imply different relationships
behaviour of electricity demand is more important now than ever be- between electricity consumption and its drivers. For example, in the
fore. eastern region the electricity consumption is mainly industry-driven
While it is relevant to conduct a study that investigates the de- whereas in western region, where we have the holy cities of Makkah
terminants of aggregate electricity demand in KSA, this is a topic that and Madinah, this is mostly population/residential driven (SAMA,
has already been addressed by the existing literature to some extent 2019). Third, the implementation of the mega projects mentioned
(Atalla and Hunt, 2016, inter alia). However, aggregate level electricity above will also have implications for electricity demand at the regional
demand has not been studied at regional level. Hence, there is a need level. Finally, recent analysis show that different regions react differ-
for a research to investigate the aggregate level electricity demand ently to the energy, of which electricity price reform, one of the key
accross regions. initiatives in the Fiscal Balance Program of the Saudi Vision 2030
The objective of this research is to understand how determinants of (Alyamani et al., 2019).
⁎
Corresponding author.
E-mail address: jeyhun.mikayilov@kapsarc.org (J.I. Mikayilov).
https://doi.org/10.1016/j.tej.2020.106772
1040-6190/ © 2020 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/BY/4.0/).
J.I. Mikayilov, et al. The Electricity Journal 33 (2020) 106772
A regional approach for KSA constitutes a novel and significant Arabia given that there are only two studies, to the best of our
contribution to the literature on electricity demand in the following knowledge, analyzing the regional electricity consumption in Saudi
ways. First, to the best of our knowledge, there are very few to none Arabia. A recent study by Alyamani et al. (2019) discussed regional
studies that examine the determinants of electricity demand at the re- aspects of electricity consumption in residential sector. However,
gional level for KSA. Alyamani et al. (2019) considered regional aspects Alyamani et al. (2019) is descriptive-type study and does not estimate
of the electricity consumption. However, they only considered re- elasticities. Diabi (1998) studied total electricity consumption for the
sidential electricity consumption and did just a descriptive analysis. five regions of Saudi Arabia, employing the data spanning from 1980 to
Diabi (1998) examined total electricity consumption for the five regions 1992. He estimated the kingdom-wise elasticities not regional-specific
of Saudi Arabia over the period 1980−1992. However, he did a panel ones. The study is relatively old which might not reflect the electricity
analysis and did not estimate region-specific parameters/elasticities, demand behaviour for the current period, due to the substantial
and hence is different from time series that we conduct in this research. changes in terms of economic development. In addition, Diabi (1998)
Additionally, he did not address integration-cointegration and other uses panel estimation techniques, without performing integration-co-
properties of the data used such as cross-sectional dependency, which integration analyses, did not address the potential cross-sectional de-
can lead a serious issue such as bias and inconsistencies and thus mis- pendency within regions. The mentioned limitations might result in
leading policy recommendations. Second, it investigates the demand for biased estimation results and consequently misleading conclusions.
electricity at a disaggregated regional level, which takes into account The studies investigating electricity consumption employed a wide
the region specific features of electricity demand behavior; Third, it range of analytical techniques that are not necessarily quantitative or
takes into account the impact of demographic factors which may play a econometrically sound. We found studies that focused on qualitative
significant role in electricity demand formation; Fourth, it uses the analysis, Granger causality analysis, simulations based on optimization
more recent data, which partially enables us to see the impact of the models, and a range of econometric estimation methods.
ongoing energy price reforms and consequences of the low oil price For example, both Hagihara (2013) (who simply described the en-
environment on electricity demand. ergy outlook in KSA) and Alrashed and Asif (2014) who conducted a
From a policy perspective, given the current transition from a survey analysis of residential electricity consumption (REC) in the
heavily subsidized electricity price environment to a market-based Eastern province of KSA can be seen as qualitative analysis.
price environment where subsidies are gradually phased out, it is im- Matar (2017) and Matar and Anwer (2017) are examples of studies
portant to consider the impact of this transition from a regional per- that adopted a simulation-based approach when investigating elec-
spective. The findings reported in this study would help a better un- tricity consumption. Both studies explored the impact of electricity
derstanding of the regional impact (on electricity demand) of different price changes in KSA on residential electricity consumption in 2011 and
price policy scenarios and changes in population and income. The 2015, respectively. In both studies, the simulations were conducted
findings will also be useful in determining which regions, and how using an optimization based partial equilibrium model. They did not
much, to provide transitional support to in order to alleviate some of explicitly consider income and demographic effects and there were no
the financial hardships associated with rising electricity prices.1 elasticities reported due to the nature of the study.
This paper employs cointegration and equilibrium error correction Since the focus of this review is on econometric studies in line with
(ECM) methodology to develop long- and short-run price, income and the nature of our research here, we continue by reviewing the existing
demographic elasticities for regional electricity demand. literature with emphasis on the type of data that was used, the
The study concludes that income, price and population are the main econometric methodology and the specifications employed, the em-
drivers of electricity demand at regional level as theoretically expected. pirical analysis strategy adopted (e.g., whether a study at hand ad-
Although, the impacts vary across regions, the estimated elasticities all dressed stochastic properties of the data) for the study.
are statistically significant, in both long and short run, and have the We compare the model specifications to the standard specification
expected signs for all the regions. The income, price and population as dictated by the theories, which requires that we account for income,
elasticities range from 0.10 to 0.93, from -0.63 to -0.06, and from 0.24 price and demographic effects (Beenstock and Dalziel, 1986; Liddle and
to 0.95, respectively, across regions in the long run. In the short run Lung, 2010; Hasanov, 2019, inter alia). This is important because the
these intervals are (0.05, 0.47), (-0.27, -0.01) and (0.13, 1.49) respec- results from studies that do not account for all factors could potentially
tively for income, price and population across the regions. The obtained contain some omitted variable bias.
Speed of Adjustment (SoA) coefficients are significant in all cases in- For industrial electricity demand in KSA, there are some studies
dicating the short-run deviations from the long-run relationship con- investigating this relationship. Al-sahlawi (1999) utilized aggregate
verge back to the equilibrium path. time series data and Eltony and Mohammad (1993) and Liddle and
The rest of the paper is structured as follows; Section 2 reviews the Lung (2010) utilized country-level panel data but did not report KSA
literature on electricity demand modeling in the case of Saudi Arabia, specific estimates. In their specifications, Al-sahlawi (1999) considered
Section 3 contains the theoretical framework and Section 4 briefly de- only income, Eltony and Mohammad (1993) considered income and
scribes the methodology used, Section 5 presents the data. Estimation price, and Liddle and Lung (2010) considered only urbanization rates.
results are presented in Section 6. In Section 7, we discuss our key More recently, Hasanov (2019) studied the determinants of industrial
findings, while Section 8 concludes the study and provides the policy electricity demand for Saudi Arabia by analyzing the core drivers of
implications where we link our findings to the current policy environ- electricity consumption in industrial sector.
ment. As mentioned earlier, the results from studies that do not account
for income, price and demographic effects might be biased. Since first
three studies mentioned above do not account for all 3 factors, they
2. Literature review may contain some bias. Furthermore, Al-sahlawi (1999) and Eltony and
Mohammad (1993) did not consider the integration-cointegration
In this section, we highlight some of the key trends in the literature properties of the variables included in their analysis before using OLS
on residential, industrial and total electricity consumption in Saudi estimation. Therefore, their results might be biased from the spurious
regression perspective. In addition, both studies are quite old and the
1
Ideally, a regional analysis for each electricity consumer category would relationship between the variables of interest might demonstrate new
provide further insights about the impact of rising electricity prices on different path with the recent data.
consumer categories and allow for a more targeted support approach, but that is Next, we turn our attention to key studies on residential electricity
beyond the scope of our analysis. Authors currently are working on this task. demand. In this area, there were some studies with a regional focus.
2
J.I. Mikayilov, et al. The Electricity Journal 33 (2020) 106772
3. Theoretical framework 2
The nominal price values were converted to real prices using GDP deflators
for the Central and Southern regions, and CPI for the Eastern and Western re-
We use a standard formulation suggested by the theories such as gions depending on which deflated price produce economically meaningful and
demand-side approach : statistically significant results.
3
J.I. Mikayilov, et al. The Electricity Journal 33 (2020) 106772
Income is proxied by GDP and disposable income (DI), both in next step and test variables for the common trend. The next section
Millions SAR, at 2010 prices. GDP data is taken from the web page of provides the results of cointegration analyses.
General Authority for Statistics of Saudi Arabia (General Authority for
Statistics (GaStat), 2018). DI data is calculated by KAPSARC re- 6.2. Cointegration test results
searchers.
Population is regional population, in persons, as a proxy for the To test the existence of cointegration relationship, i.e., whether the
demographic factors. This data was aggregated from the General variables share the long-run common trend, we employed the Engle-
Authority for Statistics of Saudi Arabia as it is available for 13 provinces Granger (EG) cointegration test and the results are given in Table 2.
(SAMA, 2019). As the table demonstrates, for all regions except the EOA, we found
Fig. 1 demonstrates the graphs of the shares of regional electricity conclusive proof of cointegration. For the three regions, all the p-values
consumption in total country level electricity consumption. from the EG tests were below the 5% level. This confirmed the presence
As can be seen from the figure the share of electricity consumption of a cointegration relationship at the 5% significance level. For further
significantly higher in the eastern region in early 90th, while decreasing investigation of the existence of the long-run relationship in the case of
over time it approached to the consumption levels in central and wes- the Eastern region, we employed the Variable Addition Test (VAT) for
tern region in recent years. cointegration, proposed by Park (1990), which states existence of co-
Theoretically, the long-run analysis can be conducted in logs or integration as a null hypothesis. The test statistic is 4.932 with the p-
levels; here we choose the log-log specification due to its ease of in- value of 0.085, concluding the cointegration relationship at 8.5 %
terpretation and compatibility with existing literature (to allow easier significance level. Considering the results of employed tests, we con-
comparisons with previous and future studies). clude that, for all the regions there is a long-run relationship.
6.1. Unit root test results After concluding the cointegration relationship among the variables
the long-run estimation results can be interpreted so they are not
As is standard practice in the literature, the main concern lies in spurious. The long-run estimation results are provided in Table 3. To
dealing with stochastic properties of the variables and hence the have an idea of the impacts in terms of magnitude ranges, we reported
Augmented Dickey-Fuller (ADF) Unit Root test was employed for each the 99 % confidence intervals for the estimated coefficients/elasticities.
variable in levels and first differences. In unit-root testing, the max- The interval estimator provides more information about the re-
imum lag length is taken to be two and optimal lag is chosen based on presentativeness quality of the estimated coefficient, in addition to
Schwarz information criterion (SIC). Table 1 displays the results of the delivering the limits of the impact.
Unit Root tests. Based on the estimation results presented in Table 3, one can see
Briefly, based on the results displayed in Table 1 we can conclude that for all the regions, the impacts of income, price and population are
that all variables are integrated of the first order [I(1)] and there is economically meaningful as they take right signs and statistically sig-
potential for cointegration relationship. Hence, we can proceed to the nificant at 1% level. As can be seen from Table 3, the income elasticity
of electricity demand varies across regions, demonstrating more stable
Table 1 behavior in the Western region and wider ranged in the Central region.
ADF Unit Root Test results. The highest income elasticity found in Central region, while it has the
smallest value in the Eastern region. Overall, the income elasticity
Region Variable (in logs) Level First Difference
ranges from 0.102 to 0.931 across the regions.
National gdp −0.948 −4.269*** Price elasticity of electricity demand is found to be ranging from
di −0.138 −5.147*** -0.607 to -0.060 across the regions.
Central dele_coa −0.976 −4.962***
The impact of population demonstrates more similar path across
pele_coa −1.274 −4.514***
pop_coa −0.271 −5.618*** regions, having differences as well. Numerically the population elasti-
Eastern dele_eoa −1.576 −4.403*** city ranges from 0.243 to 0.947.
pele_eoa −1.782 −4.465***
pop_eoa −0.582 −3.806*** 6.4. Short-run estimation results
Western dele_woa −0.903 −6.994***
pele_woa −1.898 −4.438***
pop_woa −0.139 −4.668*** For the short-run analysis, we applied Gets to ECM to determine the
Southern dele_soa −0.444 −5.110*** relationship for each region. We start with a general ECM specification,
pele_soa −1.327 −4.534*** which includes the error correction term (ECT), contemporaneous va-
pop_soa −0.670 −4.476***
lues of all independent variables and 2 lags of all the variables. We then
Note: dele = demand for electricity, COA = central operating region, exclude variables from the analysis based on the test proposed by the
EOA = eastern operating region, SOA = southern operating region, methodology and end up with the final short-run specification (e.g., see
WOA = western operating region, pop = population, pele = electricity price; Campos et al., 2005 inter alia). We are not reporting the General Un-
*** =Significant at the 1% level, intercept only case is chosen based on the restricted Models here to save space, but they are available from the
analyses. authors upon request. Table 4 documents the final ECM specification
4
J.I. Mikayilov, et al. The Electricity Journal 33 (2020) 106772
Table 2
Cointegration test results.
COA EOA SOA WOA
Notes: p-values are MacKinnon (1996) probability values; ** and *** stand for the rejection of the null of no cointegration at 5% and 1% significance level,
respectively.
Table 4
Short-run Estimation results.
Independent Variables (DLOGs) Region
ECT (−0.293, −0.151) *** (−0.892, −0.532) *** (−0.984, −0.746) *** (−1.318, −0.924) ***
income (0.152, 0.414) ** (0.054,0.182) * (0.075, 0.221) * (0.159, 0.471) *
price (−0.079, −0.013) * (−0.234, −0.088) ** (−0.274, −0.150) *** (−0.103, −0.029) *
population (0.128, 0.616) * (0.168, 0.412) ** (0.755, 1.485) *** (0.338, 0.600) ***
dele (−1) (0.118, 0.386) *
income (−1) (−0.212, −0.078) **
price (−1)
population (−1) (−0.552, 0.134) (−1.399, −0.839) ***
TESTS (p-values)
Serial Correlation LM 0.344 0.062 0.067 0.400
Normality test (Jarque-Bera) 0.850 0.856 0.522 0.300
Heteroscedasticity (White) 0.636 0.870 0.484 0.873
Ramsey RESET Test 0.471 0.153 0.133 0.942
Notes: Dependent variable is DLOG(DELE_i) and i takes COA, EOA, WOA and SOA; ***, **, * stand for rejecting a null hypothesis at 1%, 5% level and 10 % significance
levels, respectively. All independent variables except ECT are in DLOG.
5
J.I. Mikayilov, et al. The Electricity Journal 33 (2020) 106772
the lower bound of the confidence interval), the population elasticity is consumption (by transitioning to a market-based pricing scheme with
lower in the Central and Southern regions and the income elasticity is higher prices) and the negative impact of the price increase. Striking a
lower in the Eastern and Southern regions. Below we rationalize these balance will require policy makers to determine the appropriate level of
findings based on regional characteristics. household and industrial support required to dampen the negative
We suspect that the low price elasticity in the Southern region is a impact of price changes.
result of the relatively lower levels of income in the south (Household With long run elasticities that range between 0.24 and 0.95, the
Income and Expenditure Survey, GaStat, 2018). Due to the relatively strong impact of population on regional electricity demand requires
lower income levels, electricity consumption (which is mainly, about governments to pay special attention to policies that impact the total
62%, residential consumption in the region) is already conservative and and regional distribution of population, which in turn affect the ag-
optimized, and there are not many avenues to further reduce con- gregate and regional distribution of electricity demand. This is espe-
sumption as a result of price increases. This view is reinforced by the cially relevant for the effective planning of generation, transmission
relatively low income elasticity for the southern region. and distribution networks across the country. The recently im-
In order to explain the regional population elasticities, it is im- plemented expat levy is a relevant example of a policy that has the
portant to highlight that population trends are explained by migration potential to change the existing population dynamics.
patterns, births and deaths. In the Southern region where emigration is Lastly, our results show that an increase (decrease) in income (gross
prevalent due to the search for better economic opportunities in other domestic product and disposable income) leads to an increase (de-
regions by adults and population increases are mainly driven by births, crease) in electricity consumption. This result proves that electricity is a
a lower population elasticity is potentially explained by population normal good from an income perspective. It is important to note that
growth being driven by mainly births. In comparison to adults, young there are big differences in the income elasticities across regions. The
children tend to consume less electricity, hence the lower population differences range from approximately 0.7 in the COA to 0.2 in the EOA.
elasticity in the south is potentially due to population growth being This implies that it is important to account for the regional income
driven by births. The Southern and Eastern regions having lower in- differences when designing policy for the electricity sector.
come elasticities is potentially explained by the economic character- In summary, our findings constitute evidence on the characteristics
istics of the regions. In the Eastern region, which is heavily in- of the regional determinants of electricity demand in Saudi Arabia. The
dustrialized, the income elasticity being low makes sense because of the policy implications discussed above represent our contribution to the
relatively higher levels of income in the region. Changes in income are discussion in the current policy environment in Saudi Arabia.
less likely to impact the level of electricity consumption when income There is a room for future research to analyze the regional elec-
levels are high (Chang et al., 2016, inter alia). Not surprisingly, in the tricity demand based on types of consumers. This would enable pol-
Southern region where income levels are lower, the income elasticity is icymakers to see the main types by regions driving the demand for
also low. This is potentially explained by the conservative approach to electricity and clearer picture of demand behavior.
electricity consumption (of which about 62% is resudential consump-
tion in the Southern region) at relatively lower income levels. This Author contributions
makes consumers less sensitive to income changes. As Chang (1977,
1980) and Chang and Hsing (1991) discusses, demand for a particular All authors contributed equally to all aspects of the research re-
thing might be luxury for some level of income. At that level of income, ported in this paper.
depending its level, the income elasticity starts to grow rapidly, and
even might become bigger than unity. After a certain level of income, Declaration of Competing Interest
the additional increase in income does not contribute its consumption
at the same rate. Hence, the income elasticity with respect to that item The authors declare that they have no known competing financial
reduces and further increase does not change it significantly. Therefore, interests or personal relationships that could have appeared to influ-
for some low level and for higher levels of income the income elasti- ence the work reported in this paper.
cities are expected to be low. While, it can be closer, even bigger than
unity for income levels between the low and high ones. This point is Acknowledgments
also concluded by Chang and Hsing (1991) for residential electricity
consumption in US, and by Chang et al. (2016) for electricity con- The views expressed in this study are those of the authors and do not
sumption in case of panel of countries. In this regard, relatively low necessarily represent the views of their affiliated institutions. We are
income elasticities in cases of the Eastern and Southern regions can be responsible for all error and omissions.
explained by the above-mentioned points.
References
8. Conclusions and policy implications
Al-Faris, A.R.F., 2002. The demand for electricity in the GCC countries. Energy Policy 30
In this study, we examined the impact of price, income and popu- (2), 117–124.
lation (proxy for demographic effects) on electricity demand for four, Alrashed, Farajallah, Asif, Muhammad Zahid, 2014. Trends in Residential Energy
Consumption in Saudi Arabia with Particular Reference to the Eastern Province.
central, eastern, southern and western, regions in Saudi Arabia. To the Al-Sahlawi, M.A., 1999. Electricity planning with demand estimation and forecasting in
best of our knowledge, this is the first time series study to examine the Saudi Arabia. Energy Stud. Rev. 9, 82–88.
determinants of regional electricity demand in Saudi Arabia. Alyamani, Ryan, Abdulelah, Darandary, Mikayilov Jeyhun, I., Hasanov Fakhri, J., 2019.
Residential Electricity Price Reforms: Are Different Income Groups and Regions
The empirical results show the estimates of the regional long-run Impacted Equally? KAPSARC Publications. (Accessed on December 18, 2019).
price elasticities to be around −0.5 in three regions and −0.1 in the https://www.kapsarc.org/research/publications/residential-electricity-price-
Southern region. This leads to the conclusion that changes in the price reforms-are-different-income-groups-and-regions-impacted-equally/.
Atalla, T., Hunt, L., 2016. Modelling residential electricity demand in the GCC countries.
of electricity are an effective market signal that has the potential to Energy Econ. 59, 149–158.
impact the demand for electricity. Beenstock, M., Dalziel, A., 1986. The demand for energy in the UK: a general equilibrium
From a policy perspective, based on the regional differences in analysis. Energy Econ. 8 (2), 90–98.
Campos, Julia, Ericsson, Neil R., Hendry, David F., 2005. General-to-specific modeling: an
elasticities, policy makers can develop insights about the potential re-
overview and selected bibliography. FRB Int. Finance Discuss. Paper 838.
gional impact of changes in electricity prices. Furthermore, based on Chang, H.S., 1977. Functional forms and the demand for meat in the United States. Rev.
the estimated regional impact, policy makers are better positioned to Econ. Stat. 59, 355–359.
strike a balance between the desire to encourage efficient electricity Chang, H.S., 1980. Functional forms and the demand for meat in the United States: a
6
J.I. Mikayilov, et al. The Electricity Journal 33 (2020) 106772
reply. Rev. Econ. Stat. 62, 148–150. order integrated systems. Econometrica 61, 783–820.
Chang, H., Hsing, Y., 1991. The demand for residential electricity: new evidence on time-
varying elasticities. Appl. Econ. 23, 1251–1256.
Chang, Y., Choi, Y., Kim, C.S., Miller, J.I., Park, J.Y., 2016. Disentangling temporal pat- Jeyhun I. Mikayilov Jeyhun is a research fellow at
terns in elasticities: a functional coefficient panel analysis of electricity demand. KAPSARC. His primary research interests include but are
Energy Econ. 60, 232–243. not limited to applied time series econometrics, the eco-
Diabi, A., 1998. The demand for electric energy in Saudi Arabia: an empirical in- nomics of energy and environment, and sustainable devel-
vestigation. OPEC Rev. 22 (1), 13–29. opment.
Dickey, D., Fuller, W., 1981. Likelihood ratio statistics for autoregressive time series with
a unit root. Econometrica 49, 1057–1072.
Doornik, J.A., 2009. Autometrics. Chapter 4 in In: Castle, J.L., Shephard, N. (Eds.), The
Methodology and Practice of Econometrics: A Festschrift in Honour of David F.
Hendry. Oxford University Press, Oxford, pp. 88–121.
Doornik, J.A., Hendry, D.F., 2009. Modelling Dynamic Systems: PcGive 13. Timberlake
Consultants, LTD. London.
Doornik, J.A., Hendry, D.F., 2018. Empirical Econometric Modelling, PcGive 15.
Published by Timberlake Consultants Ltd Timberlake Analytics, Inc. The Loft, 2C
Blake Mews Kew Gardens Richmond, TW9 3GA, UK.
Eltony, M.N., Mohammad, Y.H., 1993. The structure of demand for electricity in the Gulf Fakhri J. Hasanov Fakhri is a research fellow leading the
Cooperation Council Countries. J. Energy Dev. 18 (2), 213–221. KAPSARC Global Energy Macroeconometric Model
Engle, R.F., Granger, C.W.J., 1987. Co-integration and error correction: representation, (KGEMM) project. His research interests and experience
estimation and testing. Econometrica 55, 251–276. span econometric modeling and forecasting, building and
General Authority for Statistics (GaStat), 2018. Household Income and Expenditure applying macroeconometric models for policy purposes,
Survey, 2018. (accessed on December 18, 2019). https://www.stats.gov.sa/en/37. energy economics with a particular focus on natural re-
Hagihara, Jun, 2013. Saudi Arabia’s domestic energy situation and policy: focusing on its source-rich countries.
power sector. Kyoto Bull. Islamic Area Stud. 6, 107–135.
Hasanov, F.J., 2019. Theoretical framework for industrial electricity consumption re-
visited: empirical analysis and projections for Saudi Arabia. KAPSARC Discussion
Paper. KAPSARC, Riyadh, Saudi Arabia. https://doi.org/10.30573/KS–2019-DP66.
Hasanov, F.J., Bulut, C., Suleymanov, E., 2017. Review of energy-growth nexus: a panel
analysis for ten Eurasian oil exporting countries. Renew. Sustain. Energy Rev. 73,
369–386.
Hendry, D.F., Johansen, S., Santos, C., 2008. Automatic selection of indicators in a fully
saturated regression. Comput. Stat. 33, 317–335 Erratum, 337-339. Waheed Olagunju Waheed is Director of Economic
Karanfil, F., Li, Y., 2015. Electricity consumption and economic growth: exploring panel- Analysis in the Policy and Strategic Planning department at
specific differences. Energy Policy 82, 264–277. Ministry of Energy in Kingdom of Saudi Arabia.
Liddle, B., Lung, S., 2010. Age structure, urbanization, and climate change in developed
countries: revisiting STIRPAT for disaggregated population and consumption-related
environmental impacts. Popul. Environ. 31 (5), 317–343.
Matar, W., 2017. A look at the response of households to time-of-use electricity pricing in
Saudi Arabia and its impact on the wider economy. Energy Strategy Rev. 16, 13–23.
Matar, W., Anwer, W., 2017. Jointly reforming the prices of industrial fuels and re-
sidential electricity in Saudi Arabia. Energy Policy 109, 747–756.
Mohammadi, H., Amin, M.D., 2015. Long-run relation and short-run dynamics in energy
consumption–output relationship: international evidence from country panels with
different growth rates. Energy Econ. 52, 118–126.
Narayan, P.K., Smyth, R., 2009. Multivariate Granger causality between electricity con-
sumption, exports and GDP: evidence from a panel of Middle Eastern countries.
Mohammad H. Al-Shehri Mohammad is head of the
Energy Policy 37 (1), 229–236.
Modeling Department at Saudi Electricity Company.
NIDLP, 2019. National Industrial Development and Logistics Program. Accessed on
Mohammad holds MS degree in Engineering.
September 11. https://vision2030.gov.sa/en/programs/NIDLP.
Park, Joon Y., 1990. Testing for unit roots and cointegration by variable addition. Adv.
Econometrics 8, 107–133 ed. by G. F. Rhodes and T. B. Fomby. Greenwich, CT: JAI
Press.
Park, J.Y., 1992. Canonical cointegrating regressions. Econometrica 60, 119–143.
Phillips, P.B., Hansen, B.E., 1990. Statistical inference in instrumental variables regres-
sion with I(1) processes. Rev. Econ. Stud. 57, 99–125.
Saikkonen, Pentti., 1992. Estimation and testing of cointegrated systems by an auto-
regressive approximation. Econ. Theory 8 (1), 1–27.
Salahuddin, M., Gow, J., Ozturk, I., 2015. Is the long-run relationship between economic
growth, electricity consumption, carbon dioxide emissions and financial development
in Gulf Cooperation Council Countries robust? Renew. Sustain. Energy Rev. 51,
317–326.
Stock, J.H., Watson, M.W., 1993. A simple estimator of cointegrating vectors in higher