Total-Factor Energy Efficiency (TFEE) and CO₂ Emissions for GCC Countries

Abstract

Recently, the potential role of energy efficiency in energy transformation on the path to sustainable development has become a crucial topic. Over the past three decades, energy consumption and CO₂ emissions in the Gulf countries have increased dramatically. This paper assesses the energy efficiency of Gulf Cooperation Council countries during the period 2000–2019. Thus, the contribution of this study to the energy policy literature is to measure the total-factor energy efficiency, in order to explore the current energy efficiency situation in the Gulf countries. This is the first study of the Gulf countries in terms of estimating the total-factor energy efficiency using the DEA–Malmquist method. The analysis shows that the average total factor productivity change index value was 0.964, with a decline rate of 3.6%. This demonstrates that energy efficiency in those regions has experienced a relative decline. The results of Malmquist analysis show that the total factor productivity change index for the Gulf countries is less than 1, which means a regression in their efficiency (energy inefficiency) from 2000 to 2019. This means that the decline in total-factor energy efficiency in those regions was mainly due to the technical progress index. The results of the study can help policy makers understand the current level of energy efficiency, and identify the main drivers of total-factor energy efficiency. Based on the results, some policy implications related to energy efficiency and suggestions for the GCC region were formulated.

1. Introduction

Governments face economic and political pressures with increasing international concern over climate change. The most important objectives for any energy policy are to improve energy efficiency. The Gulf Cooperation Council (GCC), consisting of Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates (UAE), was established based on geographic proximity, similar common objectives, and political systems. The GCC countries share a common vision for economic development plans that focus on diversification of the product base to reduce dependence on fossil fuel sources. Thus, they have a common goal to transition their economies to become less reliant on oil [1]. These countries are endowed with abundant oil and gas resources, which have been exploited, though the GCC has plans to use solar and wind energy [2]. Statistics from the International Energy Agency concerning the energy consumption trends of GCC countries from 2000 to 2019 showed an increase in their energy consumption [3]. This, in turn, affects the environment by producing carbon dioxide emissions. However, these countries in the past few decades have achieved substantial economic progress because of the massive use of hydrocarbon resources [4]. Energy efficiency and diversification have become major strategies to achieve this transition in Gulf Cooperation Council countries, since the high ratio of energy consumption (see Figure 1) in the GCC area brings environmental problems.

Economic growth relies on energy consumption, but sustainable development must be considered when setting energy policy strategies [5]. Energy efficiency has become a central focus of energy sustainability issues and national energy policies [6]. Therefore, our study concentrates on measuring energy efficiency specifically at the GCC region level; then, further environmental improvements can be planned and executed based on the research results. Using the data envelopment analysis (DEA) model and the Malmquist function, the total factor energy efficiency (TFEE) was constructed to estimate the energy efficiency of the GCC states. Gulf Cooperation Council member states (with the exception of Bahrain) have announced short- and long-term energy efficiency goals. By 2021, Saudi Arabia intended to reduce peak electricity demand by 14%, while reducing total electricity consumption by 8%. The UAE set a long-term goal of reducing energy consumption by 30 percent by 2030. In 2017, Qatar had an aggressive goal to reduce their per capita electricity consumption by 20%. Also, Kuwait set goals to improve its electricity generation efficiency by 5% by 2020 and 15% by 2030. A long-term energy strategy is currently being prepared in Oman [2].

It is crucial to improve energy consumption in the GCC region to enhance environmental quality. Thus, how will a TFEE approach with economic growth inputs and environmental output support sustainable development? Energy efficiency improves the quality of the environment, leading to sustainable development. This research used a total-factor energy efficiency approach with key economic growth inputs like labour force, capital, and energy consumption, and took into account environmental factors for Gulf Cooperation Council countries during the period 2000–2019. Our hypothesis was that assessing energy efficiency will help to improve the environmental quality in GCC countries. The contribution of this paper to the literature is that it is the first study to examine the total-factor energy efficiency for Gulf Cooperation Council countries using the DEA and Malmquist method for energy efficiency.

The remainder of the paper is organised as follows. Section 2 contains a literature review. The methodological framework of the DEA method and the Malmquist approach and the data used are presented in Section 3 and Section 4, respectively. The econometric results of the DEA–Malmquist index technique are presented in Section 5. Finally, Section 6 and Section 7 summarise the empirical results of the paper and discuss policy implications.

2. Literature Review

The past two decades have seen an increase in energy research, and energy efficiency has become a critical component of energy strategies in several countries.

Charners et al. [7] proposed the DEA–Charnes–Cooper–Rhodes (CCR) model, which is used when analysing multiple inputs and multiple outputs. Since then, several papers have been published. Many of the studies that use the DEA model are in the energy and environmental field [8]. Hu and Wang [9] first proposed the concept of total-factor energy efficiency for 29 regions in China for the period 1995–2002. They constructed a new index of energy efficiency, total-factor energy efficiency. TFEE incorporates energies, capital stock, and labour as multiple inputs to produce economic output. TFEE is a ratio of the target energy input that is suggested from data envelopment analysis (DEA) to the actual energy inputs in a region. DEA can easily be applied to a multiple input–output framework to compute the TFEE index to find regional targets of energy inputs. Thus, Hu and Wang employed labour force, fixed capital, energy consumption, and total crop area as inputs in their study, with real GDP used to estimate energy efficiency.

Caves et al. [10] first measured the change in total factor productivity using DEA and the Malmquist index, and then this was applied in many energy studies. For example, Førsund and Kittelsen [11] calculated the Malmquist productivity index for electric utilities in Norway for the period 1983–1989. They found a positive increase in productivity due to the technological shift.

Abbott [12] analysed changes in the Australian electricity industry from 1969 to 1999, using DEA and the Malmquist index to include fixed capital, labour force, and energy (in TJ) as inputs. The results showed an improvement in the industry’s performance due to the rate of improvement in productivity. Barros [13] analysed total productivity using a DEA model for hydroelectric energy generating plants in Portugal for the period 2001–2004. The study’s findings indicated that some plants had a decrease in productivity, while others experienced productivity growth. Vlahinić-Dizdarević and Šegota [14] examined energy efficiency changes at the economic level in EU countries from 2000 to 2010, using the CCR model. Their results revealed that reducing some of the inputs in the inefficient countries could improve their efficiency. Zhang et al. [15] measured total-factor energy efficiency for 23 developing countries during 1980 to 2005 using data envelopment analysis. Their results indicated that energy efficiency performance differed from country to country. However, China has achieved an increase in its total energy efficiency ratio through effective energy policies. Additional perspectives relating to research factors, such as pollutants, factor input level, and emissions, were considered in the DEA [16]. Wu et al. [17] evaluated the overall-factor energy efficiency using an envelope data analysis model for a region of China. The results showed that industrial performance improved from 2006 to 2010. Also, Liu et al. [16] constructed a combined model using data volume DEA and the Malmquist index to estimate total-factor energy efficiency (TFEE) in the thermal power industry in China from 2005–2014. They factorised the TFEE index into a technical efficiency index, a technical progress index, a net efficiency index, and a scale efficiency index. The results showed that TFEE is mainly determined by TECH and PECH. Wu et al. [18] used DEA to measure the energy and environmental efficiency of China’s industrial sector. The results showed that the sector’s efficiency is unbalanced and low.

Chen and Yang [19] measured the TFEE of Chinese provinces under resource and environmental constraints. The empirical results showed that TFEE continues to decline under constraints, and TFEE is affected by the type of structural factors. Shang et al. [20] measured the total-factor energy efficiency of various regions in China from 2005 to 2016 under environmental constraints. Their study found that the TFEE value is low, which means that energy consumption still dominates China’s economic growth. Ohene-Asare et al. [21] estimated energy efficiency using DEA for 46 African countries over the period 1980–2011. Three inputs were used to estimate TFEE: capital, labour force, and total primary energy consumed; these inputs were used to produce two outputs: real GDP and emissions. The empirical results showed that African countries achieved an average of 56% energy efficiency over the sample period. Economic development and technological advances were also found to have a significant positive impact on energy efficiency in these countries. Trotta [22] studied the impact of improvements in energy efficiency on reducing energy consumption in Finland during the period 2005–2015. The results revealed that energy efficiency saved 0.58 Mt of final energy, and the country’s carbon dioxide emissions were reduced. Recently, Shao et al. [23] examined green technology innovations’ impact on CO₂ emissions in N-11 courtiers from the period 1980 to 2018. They suggested implementing some policies to enhance renewable energy resources, as well as green innovation technologies to reduce environmental effects globally. Ren et al. [24] calculated the total-factor energy efficiency for 13 Beijing–Tianjin–Hebei cities for the period from 2007 to 2020. Their results showed that there are large differences in TFEE between the 13 cities in BTH. Zheng [25] measured energy efficiency for Shanghai using panel data for 2012–2018, in order to calculate the TFEE of each city. The overall efficiency demonstrated a downward trend; the reason behind the decline in total factor productivity was the absence of technological innovation. Recently, Li et al. [26] used a dataset consisting of 30 provinces of China from 2016 to 2019. The TFEE results showed that energy consumption on average was down by 42.36%. Thus, these regions need more government attention and resources to improve TFEE, and need to practice sustainable development. A very recent study about China reviewed the public’s perception of carbon neutrality by cluster analysis on microposts. The results showed that on the environment, social factors and governance were behind people’s concerns about carbon neutrality [27].

For GCC countries, Alsahlawi [28] attempted to examine energy efficiency in the GCC region through employing two DEA models using three inputs (capital, labour force, and energy consumption) and one output (GDP) over the period 1990–2009. The empirical results indicated that GCC countries have to improve their energy efficiency. Studies have used different methods to assess energy efficiency. Howarth et al. [29] evaluated the relationship between GDP and energy consumption at the sector level in the GCC, and found that there is strong economic growth and energy consumption in all sectors in these countries. Thus, there is a need to further improve energy efficiency in the GCC. Also, Alarenan et al. [30] used frontier analysis to measure energy efficiency for the GCC countries over the period 2004–2014 for road transport gasoline and residential electricity sectors. They found that energy efficiency in the GCC countries had generally improved. Global technological advances and stricter global fuel economy standards may have contributed to this improvement. Recently, Almasri and Narayan [4] reviewed energy efficiency and renewable energy in the GCC region. Their analysis disclosed that GCC countries’ interest in the energy efficiency field is escalating. Also, certain important measures have been undertaken in these countries to create more awareness about opportunities and goals associated with generating energy by renewable energy. They suggested that to achieve benefits from energy efficiency, a clear and easy government mechanism needs to be established. Almasri and Alshitawi [31] reviewed the electricity consumption of residential buildings in the GCC region. They found that the best energy efficiency measures involve improving the efficiency of wall insulation and air conditioners. Recently, Nikbakht et al. [32] presented an analysis of the energy efficiency and environmental performance of Persian Gulf countries in 2000–2014. Their results showed that Saudi Arabia has the highest total-factor energy efficiency, and second-highest TFEE is the United Arab Emirates, while Oman had the lowest TFEE. Also, their results implied that energy consumption could be reduced in Persian Gulf countries.

In addition, there have been studies using different methods that examined energy efficiency in separate GCC countries. Alajmi [33] found that there is a long-term relationship between CO₂ emissions and energy consumption (electricity generation) in Saudi Arabia from 1980 to 2017. The impact of energy consumption on carbon dioxide emissions will continue to be observed for decades if the country continues to use oil. Also, Alajmi [34] examined the factors influencing GHG emissions in nine sectors in Saudi Arabia. The findings showed that the energy effect in the country is the leading factor increasing emissions; thus, the country needs to increase its energy efficiency in these sectors.

The aforementioned research mainly used different methods to study energy efficiency. However, none of these studies investigated TFEE for the GCC countries using the DEA–Malmquist indices method. The GCC countries are extremely energy- and CO₂-intensive, as shown in Figure 1 and Figure 2. Therefore, it is necessary to measure the TFEE of these countries, and to know the actual situation of each region. TFEE analysis can deal with multiple inputs and outputs, and better reflects the actual situation of energy use. Since energy itself does not directly produce economic results, it must be combined with other inputs such as capital and labour to produce such results. As far as I am aware, no study has examined the overall energy efficiency of these countries.

Therefore, the contribution of this study to the energy policy literature is the application of DEA–Malmquist methods to measure energy efficiency in the GCC countries—a gap that we aim to fill. By helping policy makers identify the level of energy efficiency, we can support global efforts to improve the quality of the environment. Then, GCC countries will be able to take action to improve their energy efficiency, as part of an effective and sustainable economic development strategy.

3. Research Methods

This study evaluated the total-factor energy efficiency through a mixed DEA and Malmquist model following Liu et al. [16], Alsahlawi [28], and Hu and Wang [9]. Based on total factor productivity, we used an economic production function to analyse energy efficiencies in GCC countries. In our energy efficiency analysis, labour force, capital stock, and energy consumption were used as production inputs, and GDP and carbon dioxide emissions were used as production outputs. The DEA–Malmquist approach can consider multiple input and output indices to effectively decompose the TFEE into more specific indices. In this study, the DEA–Malmquist was employed, with and without considering carbon dioxide emissions (CO₂) as an undesired output. There were two models used in this study: the first model uses DEA–Malmquist indices for the energy efficiency with one desirable output, which is GDP, for each country. The second model has two outputs: GDP as a desirable output, and the carbon emissions as an undesirable output, for each country over the sample period.

3.1. Research Models

3.1.1. DEA Model

In this research, the total factors influencing to the energy efficiency of the energy sector were identified firstly using the data envelopment analysis (DEA) model, and then the Malmquist method. The TFEE model was constructed to assess energy efficiency of the energy sector in the GCC countries. “DEA assesses a group of decision-making units’ relative efficiencies by measuring each unit’s efficiency score using various resources (inputs) and multiple results (outputs). DEA also allows for comparison between Decision-Making Units (DMUs) and can be used in various situations” [35] (p. 3). Decision-making units (DMU) refer to GCC countries’ economies in this study. The DEA method has a wide range of applications with simple principles [36]. In this study, each DMU was set to represent decision-making units in the TFEE evaluation, using m different inputs to produce s different outputs. The efficiency was defined as the ratio of the weighted sum of outputs and the weighted sum of inputs. The Charnes–Cooper–Rhodes (CCR) efficiency model can be used to estimate the DMU [14] as follows:

$$\begin{gathered} h_0\left(u,v\right)=\frac{\sum _{r=1}^s{u}_r{y}_{r0}}{\sum _{i=1}^m{v}_i{x}_{i0}} \le 1 \\ \mathrm{s}.\mathrm{t}. \frac{\sum _{r=1}^s{u}_r{y}_{rj}}{\sum _{i=1}^m{v}_i{x}_{ij}} \le 1 (j=\mathrm{1,2},\dots ,n) \\ {u}_r \ge 0 (r=\mathrm{1,2},\dots ,s) \\ {v}_i \ge 0 (i=\mathrm{1,2},\dots ,s) \end{gathered}$$

(1)

where x stands for the mth inputs of the nth DMU, y is the sth outputs of the nth DMU. $u_r$ is the weight given to output r, $v_i$ is the weight given to input i. Weights $u_r$ i $v_i$ are chosen to maximise the efficiency of a particular DMU subject to efficiencies of all DMU in the model. If the efficiency $h_0$ is $\le 1,$ the DMU is relatively inefficient.

3.1.2. DEA–Malmquist Index

Constructed by Färe et al. (1992) from input and output data [37], the DEA–Malmquist index uses a non-parametric framework to assess the efficiency change over time using DEA technologies [38]. DEA–Malmquist approach indices have the advantages of multiple input and output indices [39]. In the aspect of measuring energy efficiency, Färe et al. [40] improved the Malmquist production index to study environmental and energy issues. In general, Malmquist total factor productivity change index (TFPCH) can be factorised into two indices, the technical progress index (TPCH) and the technical efficiency index (TECH) [16], as shown in Equation (2).

TFPCH = TPCH × TECH

(2)

$$\begin{gathered} TFPCH=\left[\frac{d_i^t\left(x_{t+1},y_{t+1}\right)}{d_i^t\left(x_t,y_t\right)}\right.\times {\left.\frac{d_i^{t+1}\left(x_{t+1},y_{t+1}\right)}{d_i^{t+1}\left(x_t,y_t\right)}\right]}^{\frac{1}{2}} \\ =\left[\frac{d_i^t(x_{t+1},y_{t+1})}{d_i^t(x_t,y_t)}\right]\times \left[\frac{d_i^t(x_{t+1},y_{t+1})}{d_i^t(x_t,y_t)}\right.\times {\left.\frac{d_i^{t+1}(x_{t+1},y_{t+1})}{d_i^{t+1}(x_t,y_t)}\right]}^{\frac{1}{2}} \end{gathered}$$

(3)

where d stands for the distance function, x as mentioned above stands for the mth inputs of the nth DMU, and y is the sth outputs of the nth DMU. When TFPCH > 1, the energy efficiency is raised from the period t to t + 1; however, when TFPCH < 1, the energy efficiency declined from the period t to t + 1. The technical efficiency index can be decomposed further into the scale efficiency index (SECH) and the pure efficiency index (PECH), as shown in Equation (4) [16].

$$TECH=\frac{d_i^{t+1}(x_{t+1},y_{t+1}\backslash C,S /d_i^{t+1}(x_{t+1},y_{t+1}\backslash V,S)}{d_i^t(x_t,y_t\backslash C,S /d_i^t(x_t,y_t\backslash V,S)}\times \left[\frac{d_i^{t+1}(x_{t+1},y_{t+1}\backslash V,S)}{d_i^t(x_t,y_t\backslash V,S)}\right]$$

(4)

In this study, total-factor energy efficiency (TFEE) can be estimated using an index of technical progress (TPCH), a pure efficiency index (PECH), and an index of scale efficiency (SECH), as shown in Equation (5).

$$\begin{gathered} TFPCH=\left[\frac{d_i^t\left(x_{t+1},y_{t+1}\right)}{d_i^t\left(x_t,y_t\right)}\right.\times {\left.\frac{d_i^{t+1}\left(x_{t+1},y_{t+1}\right)}{d_i^{t+1}\left(x_t,y_t\right)}\right]}^{\frac{1}{2}} \\ \times \frac{d_i^{t+1}(x_{t+1},y_{t+1}\backslash C,S /d_i^{t+1}(x_{t+1},y_{t+1}\backslash V,S)}{d_i^t(x_t,y_t\backslash C,S /d_i^t(x_t,y_t\backslash V,S)}\times \left[\frac{d_i^{t+1}(x_{t+1},y_{t+1}\backslash V,S)}{d_i^t(x_t,y_t\backslash V,S)}\right] \end{gathered}$$

(5)

A TFEE score greater than 1 indicates efficiency progress; however, when it equals 1, this indicates efficiency stagnation, while a score of less than 1 denotes efficiency regression (energy inefficiency) from period t to period t + 1 [41]. Coelli’s DEAP software (version 2.1) was used to solve the DEA model equations.

By the DEA–Malmquist function, the TFEE evaluation model was constructed to assess the energy efficiency of GCC countries. We set DMU to represent decision-making units in the TFEE evaluation. Thus, the TFEE is designed to overcome the disadvantages of traditional partial-factor energy efficiency indices, which consider energy as the only input to output GDP and ignore other inputs [14].

4. Data

This research empirically used a panel dataset of GCC countries over the period 2000–2019, in order to analyse energy efficiencies of these countries based on the total factor productivity method. The data for these countries are detailed in

Table 1. Research variables, data sources, sample countries, and period.

	Variable	Unit	Countries	Period	Data Source
Inputs	Labour force	Millions of people	Saudi Arabia Qatar Kuwait Oman United Arab Emirates Bahrain	2000–2019	World Bank
	Capital stock	2017 USD millions			Penn World Table
	Total final energy consumption	terajoules (TJ)			International Energy Agency
Output	Real GDP	2017 USD millions			Penn World Table
	Total CO2 emissions	Mt CO2			International Energy Agency

. In measuring the total-factor energy efficiency, mainly considering the availability of data, three input factors and two output factors were chosen to measure total-factor energy efficiency energy consumption in the GCC countries for analysis. However, in our model, the three production factors (labour force, capital stock, and energy consumption) produce two outputs, GDP as a desirable output, and CO₂ emissions as an undesirable output. The variable labour force (millions of people) was obtained from World Development indicators [42], while real GDP (2017 USD millions) and capital stock (2017 USD millions) data were obtained from Penn World Tables [43]; data for total final energy consumption in terajoules (TJ) and CO₂ emissions in millions of tonnes (Mt) were obtained from the International Energy Agency [3].

The DEA model’s summary statistics and correlation matrix of the inputs and output are presented in

Table 2. Summary statistics of input and output variables (2000–2019).

Variable	Mean	Std. Dev.	Correlation Matrix
Output			GDP	Labour	Energy	Capital	CO2 emissions
Real GDP (USD millions)	12.320	1.060	1
Input
Labour (millions)	14.527	1.017	0.95	1
Total final energy consumption (TJ)	13.559	1.081	0.97	0.97	1
Capital stock (USD millions)	13.513	1.198	0.93	0.95	0.95	1
Total CO2 emissions	13.513	0.961	0.76	0.73	0.75	0.65	1

Source: Author’s calculations.

. The correlation matrix shows that our inputs have positive correlation coefficients with the output. Positive correlation coefficients of 0.96 and 0.93 were found between the labour force and capital inputs and the GDP output, respectively. The correlation coefficient between the energy consumption input and GDP output was 0.97 as well. These results show that for the DEA model, all inputs satisfy the isotonicity property with the output.

Before presenting the DEA–Malmquist method’s results, we summarise the general economic characteristics, energy consumption, GDP, and CO₂ emissions of GCC countries over the period 2000–2019. The economies of the GCC countries still depend on oil as an energy source, and will remain so for a long time. The increase in population means greater demand for energy, putting pressure on the quality of the environment. Data on energy consumption are shown in Figure 1, which shows a continuous increase in energy consumption in all Gulf countries; however, the amount of increase in consumption depends on the country’s size and population. For example, energy consumption in Saudi Arabia and the United Arab Emirates has been continuously increasing from 2000 to 2019, with fluctuations from 2008 to 2015; however, after 2015, their energy consumption decreases slightly. For the other GCC countries, the figure shows a continued increase in energy consumption, but less than the increase seen in both Saudi Arabia and the United Arab Emirates. This may be due to differences in the size of the country and its population. In summary, Figure 1 illustrates that energy consumption increased on average from 2000 to 2019 in these countries, reflecting the dependence of these countries on fossil fuel as an energy source. The GDP is continuously increasing in all GCC countries over the period 2000–2019. Despite the increase in the contribution of non-oil sectors to the GDP, these countries are still dependent on oil revenues (see Figure A1 in Appendix A).

Figure 1. The GCC countries’ total energy consumption in units of TJ (2000–2019). Source: International Energy Agency.

Figure 1. The GCC countries’ total energy consumption in units of TJ (2000–2019). Source: International Energy Agency.

Figure 2 shows that total CO₂ emissions of the GCC region increase on average from 69.62 Mt in 2000 to 158.52 Mt in 2019. This increase coincides with the increase in energy consumption in these countries, which depend on fossil fuels as their energy source. This graph indicates that increasing energy consumption causes more carbon emissions, which in turn negatively affects the quality of the environment.

Figure 2. The GCC countries’ total CO₂ emissions by Mt (2000–2019). Source: International Energy Agency.

Figure 2. The GCC countries’ total CO₂ emissions by Mt (2000–2019). Source: International Energy Agency.

5. Empirical Results and Discussion

In this section, we assess the level of TFEE (through TFPCH) of GCC countries. Using DEAP software, the factors of production were brought into the DEA–Malmquist index. Then, the Malmquist index TFPCH, technical efficiency index (TECH), technical progress index (TPCH), pure efficiency index (PECH), and scale efficiency index (SECH) of the energy sector for GCC countries were obtained. The DEA–Malmquist indices results for the energy efficiency, with and without the carbon emissions for each country, are reported in

Table 3. The GCC countries’ TFEEs without CO₂ emissions from 2000–2019.

GCC Countries	TECH	TPCH	PECH	SECH	TFPCH
Saudi Arabia	1.004	0.967	1.000	1.004	0.972
Qatar	1.000	0.967	1.000	1.000	0.967
Kuwait	1.006	0.967	1.005	1.002	0.973
Oman	0.986	0.929	0.989	0.997	0.916
United Arab Emirates	0.986	0.988	0.995	0.991	0.974
Bahrain	1.006	0.979	1.000	1.006	0.985
Average	0.998	0.966	0.998	1.000	0.964

and

Table 4. The GCC countries’ TFEEs with CO₂ emissions from 2000–2019.

GCC Countries	TECH	TPCH	PECH	SECH	TFPCH
Saudi Arabia	1.004	0.973	1.000	1.004	0.977
Qatar	1.000	0.966	1.000	1.000	0.966
Kuwait	1.000	0.977	1.000	1.000	0.977
Oman	0.997	0.932	0.999	0.998	0.929
United Arab Emirates	0.986	0.982	0.995	0.991	0.969
Bahrain	0.998	0.984	1.000	0.998	0.983
Average	0.998	0.969	0.999	0.999	0.967

, with averages plotted in Figure 3. Estimations for the annual TFEE of all the GCC countries, without and with CO₂ emissions over the sample period, are presented in

Table 5. Total-factor energy efficiency yearly by country (2000–2019).

Year	Saudi Arabia		Qatar		Kuwait		Oman		United Arab Emirates		Bahrain
	Without CO2	With CO2	Without CO2	With CO2	Without CO2	With CO2	Without CO2	With CO2	Without CO2	With CO2	Without CO2	With CO2
2000–2001	0.961	0.961	0.906	0.904	1.032	0.978	0.941	0.929	0.842	0.842	0.952	0.966
2001–2002	0.923	0.923	0.977	0.930	1.016	1.010	0.898	0.879	0.991	0.991	0.937	0.97
2002–2003	1.065	1.065	0.869	0.874	0.978	1.062	0.884	0.861	1.031	1.031	0.955	0.978
2003–2004	1.017	1.017	0.960	1.003	1.040	1.056	0.988	0.988	1.031	1.031	0.938	1.004
2004–2005	0.987	0.987	0.932	0.930	1.003	0.989	0.799	0.804	0.951	0.951	1.009	0.937
2005–2006	0.933	0.933	0.980	1.017	1.056	1.074	0.771	0.713	1.046	1.046	1.015	0.996
2006–2007	0.888	0.888	0.942	0.956	1.035	1.076	0.938	0.890	0.877	0.877	0.971	1.025
2007–2008	1.019	1.019	0.957	0.965	0.944	0.917	0.824	0.916	0.889	0.889	0.984	0.981
2008–2009	0.946	0.933	0.945	0.956	0.940	0.892	1.041	1.033	0.828	0.826	1.018	1.038
2009–2010	1.000	0.956	1.057	1.057	0.955	0.923	0.912	0.878	0.977	0.969	0.966	0.976
2010–2011	0.983	1.002	1.052	1.052	1.011	1.043	0.975	0.901	0.992	1.024	1.013	1.007
2011–2012	1.001	0.994	0.941	0.923	0.987	1.007	0.987	1.001	0.980	1.009	1.001	0.926
2012–2013	0.980	0.975	0.990	0.990	0.976	0.961	0.941	0.953	1.013	1.027	0.969	1.013
2013–2014	1.028	1.005	0.958	0.958	0.895	0.907	0.980	0.955	1.014	1.003	1.022	0.999
2014–2015	0.998	0.945	0.981	0.979	0.931	0.902	0.939	0.916	0.991	0.960	0.985	1.005
2015–2016	0.955	0.951	0.974	0.974	0.957	0.956	0.936	0.966	1.007	0.974	0.954	1.008
2016–2017	0.959	0.967	0.977	0.977	0.869	0.829	0.990	0.990	0.979	1.056	1.006	0.963
2017–2018	0.977	0.993	0.985	0.986	0.982	0.972	0.996	0.952	1.093	1.151	0.965	0.966
2018–2019	0.960	0.960	0.986	0.957	0.987	0.972	0.954	0.942	0.918	0.912	1.017	0.963
Average	0.978	0.972	0.967	0.968	0.979	0.975	0.931	0.919	0.971	0.977	0.983	0.985

.

Table 3 shows TFEEs of the GCC countries for the study period 2000–2019; the values imply a decline in productivity growth as indicated by the TFEE value of 0.964, which shows a 3.6% decline without adding CO₂ emissions as an unwanted output. The decline in this TFPCH indicates a lower energy efficiency level in these countries. This is evident from the values of the TECH (0.998), PECH (0.998), and SECH (1.000) being stagnant. In another sense, a decline in the TFEE of GCC countries is largely due to the low level of energy technology innovation or technical progress change (TPCH) and pure efficiency (PECH). In summary, the results of the DEA–Malmquist productivity index of energy efficiency in the GCC region show a decline in productivity growth (TFPCH); the reason for the decline was also caused by technical progress change (TPCH) and the stagnation of changes in the scale efficiency index (SECH), meaning that energy efficiency has been in relative decline.

Table 4 shows that for the duration of the study (2000–2019), GCC countries displayed a decline in the TFEE value (0.967), with a 3.3% decreasing rate with CO₂ emissions as an undesired output. The results of the DEA–Malmquist productivity index of energy efficiency in the GCC region show that the reason behind the decline in productivity growth (TFPCH) was the regression in both technical progress change (TPCH) and the technical efficiency index (TECH), which was under 1, meaning that the energy efficiency experienced a relative decline.

From the results in Table 4, five countries showed a higher TFPCH than the average value: Bahrain, United Arab Emirates, Qatar, Kuwait, and Saudi Arabia. Bahrain’s productivity is higher when compared with the rest of the GCC region because efficiency change is driven by scale efficiency change. Generally during the study period, there was a decline in total productivity in all GCC countries (because they have TFPCH values of less than 1) due to technical progress TPCH.

Figure 3 shows that the average TFEE levels fluctuate during the study period, but the energy efficiency value of the GCC region without CO₂ emissions has always been higher than the energy efficiency value with CO₂ emissions during the period from 2000 to 2019. Overall, the TFEE trend seems to increase throughout the study period; the average also implies that although all input and output variables increased during the sample period, these input levels were not efficiently converted into outputs. In summary, with the TFEE analysis results (with or without CO₂ emissions as an undesirable output), the identified factors that impact the TFEE in GCC countries are technical progress change and the technical efficiency index.

Table 5 presents estimates of the annual TFEEs of all GCC countries during the sample period, with and without CO₂ emissions as an undesirable output. Table 5 shows that for Saudi Arabia, the TFPCH increased to more than 1 in five years due to the reciprocal influence of technical efficiency TECH over the period 2000–2019. For Qatar, the TFPCH exceeded 1 in four years. In Kuwait and Bahrain, the TFPCH exceeded 1 in seven years, but in Oman, the TFPCH only exceeded 1 in two years. In the United Arab Emirates, the TFPCH exceeded 1 in nine years. The analysis of the index shows that the increase in the TFPCH in the GCC countries was mainly driven by the scale efficiency index SECH. In short, the TFEE value for the GCC countries is lower than 1, which indicates efficiency regression (energy inefficiency) during the study period. Our result is in line with studies such as Liu et al. [16], Chen and Yang [19], Zheng [25], and Li et al. [26]. By contrast, our findings differ with some previous studies by Zhang et al. [15], Wu et al. [17], Nikbakht et al. [32], and Alarenan et al. [30].

The total-factor energy efficiency TFEE of GCC countries may be affected by shrinking energy demand because of the recession and the fall of oil prices in late 2008; from 2014, these effects appear to coincide with the global economic crisis. The declining TFEE for the period may also be due to decreased attention paid to energy issues during that period. However, environmental assessment systems, technical standards, and clean development plans for energy consumption should be established in the GCC region. Also, high-quality energy-saving equipment should be used for environmental protection [16]. These results imply that GCC countries need to spend more on technology, environmental policy reforms, and regulatory changes, in order to increase their energy efficiency [21].

6. Policy Implications

Although GCC member countries have announced short- and long-term energy efficiency goals, with the exception of Bahrain [2], they need to follow certain policies to raise their energy efficiency. Based on our findings, which show a decline in the total-factor energy efficiency (TFEE) value due to a decline in the technical progress index (TPCH) and the technical efficiency index (TECH), these countries need to focus on raising energy efficiency and adopting strategies to support them, such as the following: (1) Improve technologies that can contribute to energy efficiency. Also, encourage technological innovations in the energy sector that produce low fuel combustion and emissions. Investing in advanced R&D technologies is a real way to achieve energy efficiency improvements. (2) Energy efficiency strategies should be a priority, because it is necessary to use equipment and devices with energy saving and environmental protection. (3) The government should rationally reallocate industrial capital and direct the flow of resources to large installed capacity projects with high energy efficiency. Making the labour force more skilled and professional is also useful to improve energy efficiency [16]. (4) Diversify the income base in GCC countries, because they are highly dependent on oil revenues. Economic growth and energy consumption in the GCC countries are strongly linked to all sectors, meaning there is still a relationship between the GDP and domestic energy consumption [29]. (5) GCC region economics could be transferred to a green economy, which improves energy efficiency and environmental quality concurrently to enhance sustainable economic growth. (6) The Persian Gulf countries need to benefit from renewable energy more widely, and spread the culture of energy efficiency concepts. In addition, they need to help the private sector adopt renewable energy applications by issuing clear legislation and laws [4]. (7) Since the countries of the region share most of the environmental issues, such as gas emissions, they can attract foreign investment to improve service sectors through developing energy-saving and knowledge-based areas [32]. (8) The GCC countries can apply environmental policies, such as white and green certificates, which may introduce more choices for governments to follow a sustainable path to support energy efficiency strategies. Also, these policies can encourage the private sector to become involved in energy saving and renewable energy programs [44]. In addition, white certificates offer features such as reducing primary energy consumption and increasing efficiency [45]. (9) Also, the GCC countries as a group outperform their neighbors in Africa, the Middle East, as well as their non-OECD peers in most areas in terms of circular carbon economy index measures such as energy efficiency. However, these countries must do more to improve their position in the global transition to a circular carbon economy [46].

7. Conclusions

Energy efficiency is the basis of sustainable economic development, so achieving a good balance between energy consumption efficiency and economic growth may lead to sustainable development with a sufficient energy supply [9]. It is clear from the previous studies that were reviewed in this paper, especially those related to the Gulf Cooperation Council GCC countries, that the DEA–Malmquist approach was not used to estimate the total-factor energy efficiency TFEE for these countries. Therefore, the present study addressed this gap, and examined total-factor energy efficiency to explore the current level of energy efficiency.

This paper presented the results of an empirical study of energy efficiency in GCC countries based on the total-factor energy efficiency TFEE index from 2000 to 2019, using the CCR model and the Malmquist index. In our first model, three factors of production (labour force, capital stock, and energy consumption) produced one output GDP. The Malmquist analysis showed that the regression in the total-factor energy efficiency (TFEE) of energy consumption in the GCC countries was mainly due to the technical progress index (TPCH). The average value of the total factor productivity change (TFPCH) was 0.964, with a decreasing rate of 3.6%. The results of the DEA–Malmquist energy efficiency index in the GCC countries showed a decline in productivity growth (TFPCH). This decline was also shown by the technical progress index (TPCH), pure efficiency index (PECH), and stagnation in the scale efficiency index (SECH). For the second model, which had three inputs producing two outputs, GDP as a desirable output and CO₂ emissions as an undesirable output, the results show a decline in the value of total-factor energy efficiency (0.967) of 3.3%. The DEA–Malmquist productivity index findings showed that the regression in both technical progress change (TPCH) and the technical efficiency index (TECH) caused a decline in productivity growth (TFPCH).

Few energy efficiency policies have been implemented in the GCC countries. However, there is a wide range of programs that policy makers in the GCC countries are studying, such as energy efficiency labelling, minimum energy efficiency standards, and insulation regulations [30]. Thus, there is enormous potential for energy efficiency improvements in these six regions. The study results may change in the future, since the GCC countries have already increased their focus on taking actions to decarbonise the region’s power sector [47]. The results of this study can also be generalised to countries that have common characteristics with the Gulf countries, for example, Jordan.