Investigation of electricity consumption

The studies investigating electricity consumption employed a wide range of analytical techniques that are not necessarily quantitative or econometrically sound. We found studies that focused on qualitative analysis, Granger causality analysis, simulations based on optimization models, and a range of econometric estimation methods.
For example, both Hagihara (2013) (who simply described the en- ergy outlook in KSA) and Alrashed and Asif (2014) who conducted a survey analysis of residential electricity consumption (REC) in the Eastern province of KSA can be seen as qualitative analysis.
Matar (2017) and Matar and Anwer (2017) are examples of studies that adopted a simulation-based approach when investigating elec- tricity consumption. Both studies explored the impact of electricity price changes in KSA on residential electricity consumption in 2011 and 2015, respectively. In both studies, the simulations were conducted using an optimization based partial equilibrium model. They did not explicitly consider income and demographic effects and there were no elasticities reported due to the nature of the study.
Since the focus of this review is on econometric studies in line with the nature of our research here, we continue by reviewing the existing literature with emphasis on the type of data that was used, the econometric methodology and the specifications employed, the em- pirical analysis strategy adopted (e.g., whether a study at hand ad- dressed stochastic properties of the data) for the study.
We compare the model specifications to the standard specification as dictated by the theories, which requires that we account for income, price and demographic effects (Beenstock and Dalziel, 1986; Liddle and Lung, 2010; Hasanov, 2019, inter alia). This is important because the results from studies that do not account for all factors could potentially contain some omitted variable bias.
For industrial electricity demand in KSA, there are some studies investigating this relationship. Al-sahlawi (1999) utilized aggregate time series data and Eltony and Mohammad (1993) and Liddle and Lung (2010) utilized country-level panel data but did not report KSA specific estimates. In their specifications, Al-sahlawi (1999) considered only income, Eltony and Mohammad (1993) considered income and price, and Liddle and Lung (2010) considered only urbanization rates. More recently, Hasanov (2019) studied the determinants of industrial electricity demand for Saudi Arabia by analyzing the core drivers of electricity consumption in industrial sector.
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 may contain some bias. Furthermore, Al-sahlawi (1999) and Eltony and Mohammad (1993) did not consider the integration-co integration properties of the variables included in their analysis before using OLS estimation. Therefore, their results might be biased from the spurious regression perspective. In addition, both studies are quite old and the relationship between the variables of interest might demonstrate new path with the recent data.
Next, we turn our attention to key studies on residential electricity demand. In this area, there were some studies with a regional focus.
1 Ideally, a regional analysis for each electricity consumer category would provide further insights about the impact of rising electricity prices on different consumer categories and allow for a more targeted support approach, but that is beyond the scope of our analysis. Authors currently are working on this task.
J.I. Mikayilov, et al. The Electricity Journal 33 (2020) 106772
2

Precisely speaking, they focused on the Gulf Council Countries (GCC) as a region and not the regions within KSA as we do in this study. The earliest study we found was the work of Eltony and Mohammad (1993) who examined residential electricity demand for a panel of GCC countries including KSA and found long run (LR) income and price elasticities for GCC countries of 0.20 and 0.14, respectively. The study did not explicitly account for demographic effects. Al-Sahlawi (1999) estimated the short run (long run) income and price elasticity of residential electricity demand as 0.13 (0.70) and 0.10 (0.50), re- spectively. This study also did not explicitly account for demographic effects. Both studies used OLS and did not account for the integration- cointegration properties of the variables.
Atalla and Hunt (2016) use a structural time series model (STSM) and data from 1985 to 2012 to investigate residential electricity con- sumption for GCC countries. Unlike other studies examined thus far, they accounted for all three factors required to explain residential electricity consumption. They also used Cooling and Heating Degree Days variables as a proxy for weather conditions. For KSA, they found the following long-run elasticities: 0.48 for income -0.16 for price and 0.80 for population, which was the variable representing the demo- graphic effect. The study also concludes that the short-run (SR) price and population elasticities are-0.16 and 4.20, respectively, while in- come does not affect the demand in the short-run.
With regards to modelling total electricity consumption, there are several studies for KSA.
Al-Faris (2002) used a Vector Error Correction Model (VECM) ap- proach and data from 1970 to 1997 to find LR (SR) income and price elasticities for KSA of 0.05 (1.65) and -0.04 (-1.24), respectively. Narayan and Smyth (2009) using data from 1974 to 2002 and FMOLS method reported long run income elasticity of electricity consumption for KSA of -3.07. In addition, the paper does not explain/interpret the found unusual negative and substantially higher income elasticity.
Liddle and Lung (2010); Karanfil and Li (2015) and Mohammadi and Amin (2015) use ECM and panel data for many countries including KSA to investigate the causal relationship between total electricity consumption and urbanization. They found that GDP per capita and urbanization granger cause total electricity consumption per capita.
Hasanov et al. (2017) use an equilibrium correction model and panel data for a number of oil exporting countries including KSA to investigate the relationship between GDP, residential electricity con- sumption, foreign direct investment and employment. One of their key findings was that employment Granger-causes residential electricity consumption in the short run. Similarly, Salahuddin et al. (2015) using panel data for GCC countries found that GDP per capita granger causes total electricity consumption per capita and reported a LR income elasticity of 0.41.
Diabi (1998) analyzed regional total electricity consumption in KSA, based on panel data (19801992) for five KSA regions at the time (CR, WR, ER, SR and NR). The study compared the results of different esti- mation methods (OLS, CHTA, CCTA, FE, RE and MLE) and accounted for income, price and demographic effects using the urbanization rate. He reported long run elasticities for KSA of 0.09 to 0.49 for income, 0.14 to 0.00 for price and 0.931.30 for urbanization. Corresponding SR elasticities are income (0.05 to 0.33), price (-0.12 to 0.00) and ur- banization (0.621.10).
In summary, there are a number of studies examining residential, industrial and total electricity consumption in KSA, but to the best of our survey, there are no studies that examine regional total electricity consumption. Considering this fact, the current study aims to in- vestigate the determinants of electricity demand for Saudi Arabia at the regional level using different cointegration techniques.
3. Theoretical framework
We use a standard formulation suggested by the theories such as demand-side approach :
Electricity use = F (price, income, population)
Where Electricity use is the total regional electricity demand, Income is income proxy, Population is population size of the appropriate region and Price is real electricity price.
The econometric functional relationship can be formulated as follow:
= + + + +Electricity use Income Population Price e0 1 2 3
Since increases in income and population increase electricity de- mand, while price increases negatively affect the demand of electricity, the expected signs for the coefficients , and 1 2 are positive while 3 it expected to be negative. e
All the variables are in logarithmic form in above specification; hence, the coefficients are elasticities, which capture the percentage change in electricity use as a result of a 1% change in the variable consideration

PLACE THIS ORDER OR A SIMILAR ORDER WITH ASSIGNMENT GURUH TODAY AND GET AN AMAZING DISCOUNT