Unpacking How Natural Gas, Digital Growth, and Hydro-Based Energy Sources Impact Ecological Sustainability in Egypt

Abstract

Egypt, as a nation, is committed to achieving ecological sustainability, which helps to protect the environment for future generations, thereby ensuring a balance between energy production, environmental health, and economic development. With regard to this vision, this research utilized the autoregressive distributed lag (ARDL) estimator to probe how hydroelectricity, digitalization, and natural gas affect ecological degradation within the Environmental Kuznets Curve (EKC) in Egypt. This study further used two distinct environmental proxies, namely, CO₂ emissions and the ecological footprint. The result of the ARDL estimator indicates that there is an inverted U-shaped association between economic growth and environmental sustainability, while urbanization does not affect environmental sustainability. Moreover, hydroelectricity, digitalization, and natural gas negatively impact environmental sustainability in Egypt. Furthermore, the frequency domain causality approach showed that there is a two-way causality pathway between environmental sustainability and its regressors. Based on this outcome, policymakers should prioritize decoupling economic growth from environmental degradation by investing in green infrastructure, sustainable industries, and circular economy models.

1. Introduction

Achieving ecological sustainability has become a critical goal for countries globally, particularly as they confront the escalating impacts of climate change and environmental degradation, which threaten both natural ecosystems and human livelihoods [1,2,3]. The urgency of this goal is underscored by the increasing frequency of extreme weather events, rising sea levels, and resource scarcity that jeopardize food and water security [4]. Moreover, Prior studies [3,4,5,6] concluded that these environmental issues are linked to the energy derived from fossil fuel-based energy sources. For Egypt, the challenge of balancing rapid economic growth with ecological sustainability is particularly pronounced, given its significant energy demands and reliance on various energy sources, including fossil fuels and renewables [7,8]. As one of the most populous countries in the Middle East and North Africa (MENA) region [9], Egypt faces mounting pressure to meet the energy needs of its growing population while simultaneously striving to reduce ecological degradation [10]. Moreover, the Egyptian participation in COP28 highlighted its commitment to addressing climate change impacts and promoting a flexible, environmentally friendly future. For effective environmental strategies, it is essential to thoroughly investigate the key drivers of ecological sustainability, ensuring that policies are both well-informed and capable of addressing the underlying challenges.

Among many other drivers, digitalization (DIGIT) has been pinpointed as a main factor in the global pursuit of ecological sustainability. DIGIT has rapidly transformed various sectors globally, offering innovative solutions to enhance efficiency and reduce resource consumption [11,12]. Lee et al. [13] emphasized that digital technologies have been increasingly promoted to facilitate economic development, enhance public services, and optimize energy usage. As a result, Ullah et al. [14] concluded that DIGIT could be detrimental to ecological quality, and thereby managing the potential of DIGIT is vital. Moreover, Adeshola et al. [15] stressed that DIGIT has the potential to contribute to ecological sustainability, as it can improve monitoring systems, promote resource efficiency, and encourage the adoption of green technologies. Therefore, understanding the implications of DIGIT for ecological sustainability in the Egyptian context requires to be investigated.

Hydroelectricity (HE) has long been a cornerstone of Egypt’s energy strategy, primarily through the Aswan High Dam, which generates a significant portion of the country’s electricity [16,17]. As a renewable energy source, HE presents an opportunity to reduce reliance on fossil fuels and mitigate environmental degradation [18,19]. However, [20] emphasized that large-scale HE projects can also impose ecological costs, such as alterations to aquatic ecosystems and the displacement of local communities. Therefore, a comprehensive assessment of HE’s role in Egypt’s energy landscape is essential to determine its overall sustainability and impact on the environment.

Natural gas (NGAS) is increasingly central to Egypt’s energy security, particularly following the discovery of substantial reserves in the Eastern Mediterranean [21]. While NGAS serves as a cleaner alternative to more carbon-intensive fossil fuels like coal and oil, its classification as a fossil fuel necessitates a careful evaluation of its long-term sustainability [22]. As Egypt expands its NGAS infrastructure to meet growing energy demands, understanding the role of natural gas in reducing the country’s carbon footprint while balancing economic growth is critical.

This study aims to analyze the role of DIGIT, HE, and NGAS in shaping ecological sustainability within the framework of the Environmental Kuznets Curve (EKC) hypothesis in Egypt. By examining these relationships, the research seeks to address the following key questions:

• How does digitalization impact ecological sustainability in Egypt?
• What role does hydroelectricity play in mitigating ecological degradation, and what are the environmental trade-offs associated with its use?
• Is the EKC hypothesis valid in Egypt?
• To what extent can natural gas function as an effective transitional energy source for reducing carbon emissions, and how does its ecological impact compare to that of hydroelectricity?

The contribution of this research lies as the first study to examine how DIGIT interacts with traditional and transitional energy sources to influence environmental outcomes in Egypt. While previous research has focused on the environmental impacts of renewable energy or technological innovation in isolation, this study explores the synergy between DIGIT, HE, and NGAS as a major force in driving Egypt’s progress toward ecological sustainability. Additionally, the study also focuses on urbanization as a controlling factor in energy use, and environmental management is explored in the context of Egypt. As digitalization gains momentum globally, understanding its potential to enhance Egypt’s ecological quality while supporting industrial and economic growth is crucial.

Secondly, this study utilized two distinct environmental proxies, CO₂ emissions and the ecological footprint, to provide a more comprehensive analysis of how various factors impact Egypt’s ecological sustainability. By examining both proxies, the study captures the multidimensional aspects of environmental degradation, in which CO₂ emissions reflect direct contributions to climate change through the release of greenhouse gases, while the ecological footprint accounts for a broader range of environmental pressures, such as land use, resource consumption, and waste generation. This approach clearly provides insights into the specific environmental impacts associated with each factor in Egypt. By employing both metrics, the study provides a richer, more detailed picture of Egypt’s ecological challenges and the pathways available for mitigating its environmental footprint.

Thirdly, the choice of Egypt as the focus of this study is grounded in the country’s unique position as a regional leader in renewable energy and digital transformation. Egypt’s growing role in international climate negotiations, as well as its domestic efforts to diversify its energy mix, make it an important case study for assessing the effectiveness of sustainability strategies in emerging economies. The country’s ambitious plans to scale up renewable energy projects, such as the development of large-scale solar and wind farms, reflect its commitment to reducing carbon emissions while meeting the energy demands of its rapidly growing population. Furthermore, Egypt’s abundant natural gas reserves have positioned it as a key player in the global energy market, but the ecological implications of its reliance on natural gas require closer scrutiny. This study seeks to provide a comprehensive analysis of how Egypt can balance its energy needs with its environmental responsibilities, offering policy recommendations that align with both national and global sustainability targets.

Fourthly, the findings are expected to provide valuable insights for policymakers, energy stakeholders, and environmental advocates, helping to chart a path toward a more sustainable and resilient future for Egypt. Given the increasing urgency of addressing climate change and the need for innovative solutions to mitigate its impacts, this study aims to bridge the gap between technological advancements and energy policy, offering a holistic approach to environmental management that is both forward-looking and grounded in Egypt’s unique energy context.

The structure of this study is as follows: an overview of related studies is provided in Section 2, while Section 3 presents details on the dataset and methodology. Subsequently, Section 4 delves into a detailed presentation and discussion of the findings, followed by Section 5, which concludes the research by presenting policy inferences and recommends pathways for subsequent studies.

2. Literature Review

Identifying the interconnections between several factors affecting environmental degradation is essential for formulating effective methods to improve ecological sustainability. This literature review will analyze the relationship among these aspects within the theoretical framework, offering a thorough understanding of the subject and highlighting the gaps in the literature. This section is categorized into four sections, namely: (i) digitalization (DIGIT) and ecological quality, (ii) economic growth and ecological quality, (iii) Natural gas (NGAS) and ecological quality, (iv) urbanization and ecological quality, and (v) hydroelectricity (HE) and ecological quality.

2.1. Digitalization and Ecological Quality

Adeshola et al. [15] suggested that the investment and development in technological innovation toward the digital flow of information will heighten research and development cooperation both inside and across businesses. Furthermore, prior studies such as [23,24,25] have confirmed the positive effect of DIGIT on economic expansion and productivity. Due to their role toward productivity, DIGIT could improve or impede environmental quality. For instance, Mehmood et al. [26] studied the role of DIGIT on ecological quality in G8 nations, while including green energy, financial development, economic growth, and technological innovation. Their result confirmed the positive role of green energy, financial development, and technological innovation on ecological quality, while the negative effect of economic growth on ecological quality was observed. Likewise, a study conducted on the E7 nations by Zhang et al. [27] examined the effect of DIGIT on ecological sustainability, including other factors such as renewable energy, financial development, economic growth, and population growth. The authors found that financial development, economic growth, and population growth impede ecological sustainability, while renewable energy and DIGIT improve ecological sustainability in E7 nations. Additionally, the work of Yao et al. [28] employed two different models to uncover the negative effect of industrial DIGIT on CO₂ emissions in 40 nations. Meanwhile, Ullah et al. [14] observed the positive impact of DIGIT on CO₂ emissions in OECD nations. As a result, the authors argued that there is a need to establish a balance in managing the potential of DIGIT across the studied nations. Meanwhile, Ha et al. [29] studied the effect of DIGIT as a driver of ecological performance in European nations and observed its negative role in the short term, while its positive role is evident in the long term. Likewise, Škare et al. [30] discovered the negative effect of DIGIT on CO₂ emissions in European nations. From the empirical literature, we observed that the nexus between DIGIT and ecological sustainability is inconsistent, and the issue remains unresolved.

2.2. Economic Growth and Ecological Quality

Economic expansion induces a significant increase in the use of resources for production processes, which directly impacts ecological quality [31,32,33]. As a result, several prior empirical studies have been conducted regarding the nexus between economic growth (GDP) and the environment for different nation(s) and sets of nations. For instance, Ullah et al. [14] confirmed that there is a N-shaped association between GDP and CO₂ emissions in OECD nations, while financial inclusion reduces CO₂ emissions. Moreover, Udemba et al. [34] further explained the nexus between GDP and ecological footprint (EF) in BRICS nations. Damak and Eweade [35] investigated the effect of GDP on EF in South Korea and uncovered that GDP positively impacts EF, while renewable energy and the rule of law negatively impact EF. Meanwhile, Somoye and Ayobamiji [4] probed the role of GDP on EF and CO₂ emissions in BRICS nations. The result confirmed that GDP and forest rent increase EF and CO₂ emissions, while renewable energy and financial development reduce EF and CO₂ emissions. Boluk and Karaman [36] studied the role of GDP on EF in Türkiye. The authors’ empirical result indicated that the EKC is valid, while agriculture induces EF. Daştan and Eygü [37] also confirmed the validity of the EKC in Turkey, while foreign direct investment and renewable energy reduce EF. Likewise, the study of Mohamed et al. [38] also corroborated this result by establishing the validity of the EKC in Malaysia. Moreover, the work of Shah and Ximei [39] also verified the presence of EKC conducted for BRICST nations and discovered that green technological innovation and renewable energy reduce EF. Likewise, in a study conducted for N-11 nations by Usman et al. [40], the validity of the EKC hypothesis was also uncovered. Based on the empirical literature, testing the validity of the EKC has been inconclusive, and this concern remains unresolved.

2.3. Natural Gas (NGAS) and Ecological Quality

Energy consumption is a vital input for production and industrial processes, which directly impact the environment. However, Ding et al. [41] argued that the detrimental impact of energy consumption varies depending on the energy source. As a result, prior studies have explored the effect of energy sources, particularly NGAS, on ecological performance. For instance, Dong et al. [42] explored the role of NGAS on CO₂ emissions in China. Their result confirmed that NGAS and renewable energy reduce CO₂ emissions, while the positive effect of economic growth on CO₂ emissions was observed. Likewise, Shah et al. [43] revealed that both the moderating and main role of NGAS has a reducing effect on CO₂ emissions in 13 selected nations. Meanwhile, the work of Azam et al. [44] also explored the role of NGAS on CO₂ emissions in 10 selected nations, but their result revealed that NGAS induces CO₂ emissions. Likewise, the detrimental influence of NGAS on ecological sustainability in India was observed by the work of Adebayo et al. [45]. Additionally, Zeraib et al. [46] studied the role of NGAS on carbon footprint in France and discovered the positive role of NGAS toward carbon footprint. Based on the empirical literature, investigating the nexus between NGAS and the environment has been inconclusive, and this concern remains unresolved.

2.4. Urbanization (URB) and Ecological Quality

Usman et al. [40] studied the effect of URB on ecological sustainability in N-11 nations as well as other factors such as GDP and financial development. The results showed that URB and financial development contribute to ecological sustainability. Likewise, Abdullahi et al. [47] suggested that URB negatively impacts EF in ECOWAS nations, while financial development and trade openness increase EF. Meanwhile, Udemba et al. [34] confirmed the positive role of URB on EF in BRICS nations. Likewise, Ntom Udemba et al. [48] further revealed that URB positively impacts EF in Russia. Mehmood et al. [49] investigated how URB affects EF in G-11 nations and uncovered that URB positively impacts EF, while eco-innovation and globalization negatively impact EF. Tanveer et al. [50] looked into how URB impacts environmental degradation in Pakistan. The result confirmed that URB increases environmental degradation. Conversely, Adebayo and Ullah [51] studied the role of URB on CO₂ emissions in Sweden. Their result indicated that URB has a negative correlation with CO₂ emissions in Sweden. Likewise, the study of Anser et al. [22] and Ramzan et al. [52] also corroborated this result by establishing the negative impact of URB on CO₂ emissions in Finland and China, respectively. Meanwhile, Ullah and Lin [53] revealed that URB increases EF in Pakistan. Based on the empirical literature, inspecting the nexus between URB and environmental sustainability has been inconclusive, and this concern remains unresolved.

2.5. Hydroelectricity (HE) and Ecological Quality

Lau et al. [54] studied the causality linkage between HE and CO₂ emissions in Malaysia, which revealed a one-way causality association between HE and CO₂ emissions. Meanwhile, Özbay et al. [55] investigated the impact of HE on CO₂ emissions in China. Their results showed that HE reduces CO₂ emissions in China. Likewise, the work of Saadaoui et al. [56] also discovered that HE decreases CO₂ emissions in Turkey. Zheng et al. [57] examined the effect of HE on CO₂ emissions in Bangladesh and factors, such as economic globalization and GDP. The authors found that HE and economic globalization impede CO₂ emissions, while GDP induces CO₂ emissions in Bangladesh. Xiaosan et al. [58] assessed the impact of HE on CO₂ emissions in China. Their result confirmed that HE and green innovation reduce CO₂ emissions, while the positive effect of GDP and foreign direct investment on CO₂ emissions was observed. Likewise, Akadiri et al. [59] revealed the mitigating role of hydro energy on EF in India.

The review of prior literature showed significant gaps that merit attention. Environmental assessments of energy consumption are frequently confined to comparisons between overall renewable and non-renewable sources, with limited focus on the implications of natural gas and hydroelectricity. Furthermore, the role of digitalization in influencing environmental quality remains a nascent and evolving area of inquiry, particularly within the context of Egypt. This study aims to address these critical gaps, contributing to a more comprehensive understanding of the interplay between natural gas, hydroelectricity, and environmental outcomes amidst digital growth and urban transformations in Egypt.

3. Data and Methods

This study focuses on investigating how DIGIT, HE, NGAS, and URB affect ecological sustainability within the EKC framework. Moreover, this research used a dataset from Egypt spanning between 1961 and 2021. In this point, the study considers two sub-types of ecological sustainability, namely CO₂ emissions and ecological footprint. Data on CO₂ emissions and ecological footprint were obtained from the BP Global (2023) database and Global Footprint Network (2023), respectively. The main explanatory variables for these dependent variables are digitalization, the square of economic growth (GDPSQ), hydroelectricity, natural gas, urbanization, and economic growth. Data on digitalization, economic growth, and urbanization were obtained from the World Bank (2023) database, while the dataset of hydroelectricity and natural gas was obtained from the BP Global (2023) database.

Table 1. Variables information.

Indicators	Codes	Metric	Sources
CO2 emissions	CO2	Million tons	[60]
Ecological footprint	EF	Global hectares per capita	[61]
Urbanization	URB	% of total population	[62]
Digitalization	DIGIT	Fixed telephone subscriptions (per 100 people)
Natural gas	NGAS	Exajoules	[60]
Hydroelectricity	HE	Exajoules
Economic growth	GDP	GDP per capita (constant 2015 US$)	[62]
Square of GDP	GDPSQ	GDP per capita (constant 2015 US$)

explains the variables.

3.1. Theoretical Framework and Model Specification

The study’s model reflects both the IPAT framework and the Environmental Kuznets Curve (EKC) hypothesis. Holdren and Ehrlich [63] developed the IPAT to explain how human activities affect the environment. Additionally, the improvement of the IPAT by Dietz and Rosa [64] to include the Stochastic Impacts by Regression on Population, Affluence, and Technology (STIRPAT), which captures the environmental impact as a function of population, affluence, and technology and expressed in a mathematical form in Equation (1) as follows:

$$I=C. P_i^{\alpha }. A_i^{\beta }.T_i^{\gamma }.{\epsilon }_i$$

(1)

where I indicates environmental impact, A indicates affluence, and T and P indicate technology and population. For this study, environmental sustainability (ENVSU) depicts the environmental impact, affluence depicts GDP per capita, technology indicates digitalization, and population is depicted as urbanization, which is expressed in Equation (2) as follows:

$${ENVSU}_t=f\left({GDP}_t, {DIGIT}_t, {URB}_t\right)$$

(2)

The EKC hypothesis highlights the inverted U-shaped relationship between environmental sustainability and per capita income. However, testing the hypothesis remains contested in light of recent studies on the EKC. This study tested whether the EKC is present in Egypt by building upon previous research such as [56,57,58], which is expressed in Equation (3) as follows:

$${ENVSU}_t=f\left({GDP}_t,{GDPSQ}_t, {DIGIT}_t, {URB}_t\right)$$

(3)

Furthermore, this study further expands the ENVSU function by including hydroelectricity and natural gas, which aligns with prior studies such as [34,36,37,45,65,66,67]. The ENVSU function is modified in Equation (4) as follows:

$${ENVSU}_t=f\left({GDP}_t,{GDPSQ}_t, {HE}_t,{NGAS}_t,{DIGIT}_t, {URB}_t\right)$$

(4)

As mentioned earlier, ENVSU is sub-divided into the CO₂ emissions function and EF function, which is expressed in Equations (5) and (6) as follows:

$$\mathrm{Model} 1: {CO}_{2t}=f\left({GDP}_t,{GDPSQ}_t, {HE}_t,{NGAS}_t,{DIGIT}_t, {URB}_t\right)$$

(5)

$$\mathrm{Model} 2: {EF}_t=f\left({GDP}_t,{GDPSQ}_t, {HE}_t,{NGAS}_t,{DIGIT}_t, {URB}_t\right)$$

(6)

where CO₂ indicates CO₂ emissions, and EF depicts ecological footprint. GDP depicts GDP per capita. GDPSQ, DIGIT, URB, HE, and NGAS indicate the square of GDP per capita, digitalization, urbanization, hydroelectricity, and natural gas, respectively. Subscript t depicts the period of study (1961–2021). Additionally, a logarithmic transformation for each dataset was conducted so as to reduce the variability of the data.

3.2. Econometric Strategy

This study used the autoregressive distributed lag (ARDL) estimator to inspect the long-term effect of DIGIT, HE, NGAS, and URB on ecological sustainability in Egypt. According to Pesaran et al. [68], there are two requirements necessary before analyzing the ARDL estimator on the dataset. The first condition requires that the explained variables are only stationary at first difference I(1), while the second requirement necessitates a long-term association between the explained variables and their explanatory variables. This study investigated whether the studied variables are stationary either at I(1) or I(0) by using the two unit root testing approach, namely the Augmented Dickey–Fuller (ADF) unit root test and the Phillips–Perron (PP) unit root test suggested by [69,70], respectively. These tests are crucial for mitigating misleading regression effects associated with non-stationary features, thereby enhancing the model’s stability and resilience. Upon verifying that the dependent variable (CO₂ and EF) exhibits stationarity at I(1), the optimum lag for the model is determined using the Schwarz criterion (SIC). Furthermore, this study used the bound test of [68] and integrated the critical values of [71] to conduct the cointegration association between the explained variables and their regressors.

Upon verifying the stationarity and cointegration assumptions, it is necessary to estimate the ARDL model for both the long run and short run. The ARDL model effectively captures the estimates for the long-run and the short-run coefficient (β), which allows for a detailed examination of their association. The bound test model used to evaluate the long-term associations among the research variables is delineated in Equations (7) and (8) as follows:

$$\begin{array}{l}∆{CO}_{2t}={\beta }_0+{\beta }_1{CO}_{2_{t-1}}+{\beta }_2{GDP}_{t-1}+{\beta }_3{GDPSQ}_{t-1}+{\beta }_4{HE}_{t-1}+{\beta }_5{NGAS}_{t-1}+{\beta }_6{DIGIT}_{t-1}+{\beta }_7{URB}_{t-1}+{\displaystyle \sum _{i=0}^m}{\theta }_1{∆{CO}_2}_{t-i}+{\displaystyle \sum _{i=0}^n}{\theta }_2{GDP}_{t-1}\\ +{\displaystyle \sum _{i=0}^p}{\theta }_3{GDPSQ}_{t-1}+{\displaystyle \sum _{i=0}^x}{\theta }_4{HE}_{t-1}+{\displaystyle \sum _{i=0}^k}{\theta }_5{NGAS}_{t-1}+{\displaystyle \sum _{i=0}^r}{\theta }_6{DIGIT}_{t-1}+{\displaystyle \sum _{i=0}^s}{\theta }_7{URB}_{t-1}\\ +{\epsilon }_t\end{array}$$

(7)

$$\begin{array}{c}∆{EF}_t={\beta }_0+{\beta }_1{EF}_{t-1}+{\beta }_2{GDP}_{t-1}+{\beta }_3{GDPSQ}_{t-1}+{\beta }_4{HE}_{t-1}+{\beta }_5{NGAS}_{t-1}+{\beta }_6{DIGIT}_{t-1}+{\beta }_7{URB}_{t-1}+{\displaystyle \sum _{i=0}^m}{\theta }_1{∆EF}_{t-i}+{\displaystyle \sum _{i=0}^n}{\theta }_2{GDP}_{t-1}\\ +{\displaystyle \sum _{i=0}^p}{\theta }_3{GDPSQ}_{t-1}+{\displaystyle \sum _{i=0}^x}{\theta }_4{HE}_{t-1}+{\displaystyle \sum _{i=0}^k}{\theta }_5{NGAS}_{t-1}+{\displaystyle \sum _{i=0}^r}{\theta }_6{DIGIT}_{t-1}+{\displaystyle \sum _{i=0}^s}{\theta }_7{URB}_{t-1}+{\epsilon }_t\end{array}$$

(8)

where ∆ signifies the first-difference operator, which captures the temporal variation in the variables. Subscript of t − 1 indicates the lag length, which is based on SIC, and the error term is depicted as ${\epsilon }_t$. The null hypothesis of the test posits the absence of cointegration among the variables, while the alternative hypothesis asserts the presence of cointegration between the variables. After identifying the long-run equilibrium connections among the research variables, the ARDL estimator is used to ascertain the short and long-run coefficients. Moreover, the error correction test of the bound test is the equation used for executing the ARDL estimator, which is articulated in in Equations (9) and (10) as follows:

$$\begin{array}{c}∆{CO}_{2t}={\beta }_0+{\beta }_1{CO}_{2_{t-1}}+{\beta }_2{GDP}_{t-1}+{\beta }_3{GDPSQ}_{t-1}+{\beta }_4{HE}_{t-1}+{\beta }_5{NGAS}_{t-1}+{\beta }_6{DIGIT}_{t-1}+{\beta }_7{URB}_{t-1}+{\displaystyle \sum _{i=0}^m}{\theta }_1{∆{CO}_2}_{t-i}+{\displaystyle \sum _{i=0}^n}{\theta }_2{GDP}_{t-1}\\ +{\displaystyle \sum _{i=0}^p}{\theta }_3{GDPSQ}_{t-1}+{\displaystyle \sum _{i=0}^x}{\theta }_4{HE}_{t-1}+{\displaystyle \sum _{i=0}^k}{\theta }_5{NGAS}_{t-1}+{\displaystyle \sum _{i=0}^r}{\theta }_6{DIGIT}_{t-1}+{\displaystyle \sum _{i=0}^s}{\theta }_7{URB}_{t-1}+{\theta ECT}_{t-1}+{\epsilon }_t\end{array}$$

(9)

$$\begin{array}{c}∆{EF}_t={\beta }_0+{\beta }_1{EF}_{t-1}+{\beta }_2{GDP}_{t-1}+{\beta }_3{GDPSQ}_{t-1}+{\beta }_4{HE}_{t-1}+{\beta }_5{NGAS}_{t-1}+{\beta }_6{DIGIT}_{t-1}+{\beta }_7{URB}_{t-1}+{\displaystyle \sum _{i=0}^m}{\theta }_1{∆EF}_{t-i}+{\displaystyle \sum _{i=0}^n}{\theta }_2{GDP}_{t-1}\\ +{\displaystyle \sum _{i=0}^p}{\theta }_3{GDPSQ}_{t-1}+{\displaystyle \sum _{i=0}^x}{\theta }_4{HE}_{t-1}+{\displaystyle \sum _{i=0}^k}{\theta }_5{NGAS}_{t-1}+{\displaystyle \sum _{i=0}^r}{\theta }_6{DIGIT}_{t-1}+{\displaystyle \sum _{i=0}^s}{\theta }_7{URB}_{t-1}+{\theta ECT}_{t-1}+{\epsilon }_t\end{array}$$

(10)

where ECT depicts the error term correction, indicating the rate at which the model reverts to its long-term equilibrium after experiencing short-period disruptions, while $\theta$ depicts the coefficient of the ECT. Finally, we investigate the causality association between ENVSU and its regressors by using the frequency domain causality approach developed by [72]. This approach offers benefits over conventional time-domain procedures by allowing more variability than regular time-domain tests, which are limited to fluctuation within a fixed period. The frequency domain technique exhibits greater resilience to seasonal fluctuations. Furthermore, the methodology remains consistent in nonlinear topography, with causation cycles manifesting at elevated or diminished frequency. For statistical analysis, this study used a series of software such as Eviews 12 and Stata 15.

4. Presentation of Result and Discussion

Presentation of Result

The research begins with inspecting the essential attributes, which are presented in

Table 2. Descriptive statistics.

	EF	CO2	DIGIT	GDP	GDPSQ	HE	NGAS	URB
Mean	0.104	1.907	6.265	3.294	10.896	0.966	1.659	1.632
Median	0.124	1.957	6.349	3.309	10.953	1.020	2.036	1.632
Maximum	0.248	2.348	7.073	3.590	12.893	1.178	2.793	1.642
Minimum	−0.133	1.200	5.343	2.954	8.726	0.238	−0.326	1.598
Std. dev.	0.112	0.349	0.627	0.201	1.314	0.225	1.074	0.009
Skewness	−0.633	−0.508	−0.179	−0.322	−0.252	−1.872	−0.819	−1.674
Kurtosis	2.354	2.026	1.434	1.860	1.832	5.966	2.229	5.953

. We observed that GDPSQ has the highest average value and standard deviation among the dataset, followed by DIGIT. Meanwhile, EF has the lowest average value, while URB exhibits the least variation value. Furthermore, the variation of each variable falls within the mean value except for EF. Furthermore, all variables are negatively skewed. The peakness of each variable show that it is platykurtic (less than 3) except for HE and URB.

Furthermore, the results of the ADF and PP unit root tests are presented in

Table 3. ADF and PP tests.

	ADF		PP
	Level	Δ	Level	Δ
ECF	−0.982	−7.143 *	−1.133	−7.188 *
CO2	−0.960	−8.112 *	−0.960	−8.099 *
GDP	−0.861	−4.702 *	−1.473	−4.305 *
GDPSQ	−2.067	−4.642 *	−1.728	−4.329 *
DIGIT	−1.842	−4.192 *	−1.249	−4.186
EGLO	−2.903	−7.794 *	−3.216	−7.913 *
HE	−4.595 *	−4.385 *	−4.146 *	−4.336 *
NGAS	−1.310	−3.232 ***	−0.978	−7.274 *
URB	−3.776 **	−1.735	−4.028 **	−1.164

Note: 10%, 5% and 1% level of significance denotes ***, ** and *, respectively.

. The result of both tests revealed that there is a unit root issue (i.e., non-stationary) for all variables at level (I(0)) except for HE and URB. Upon differentiating these variables, the results showed that the remaining variables have no unit root issue (i.e., stationary). Thus, it shows that there is a mixed order of integration, which meets the requisite requirements for the application of the ARDL model.

For the cointegration analysis,

Table 4. Bound cointegration test outcome.

	Model 1	Model 2
F-stat	9.438 *	9.507 *
T-stat	−8.653 *	−8.913 *
Diagnostic check
	Model 1	Model 2
χ2 Normality	1.622 (0.444)	4.512 (0.104)
χ2 LM	0.300 (0.742)	1.900 (0.139)
ARCH heteroscedasticity	2.251 (0.139)	2.671 (0.109)
χ2 Ramsey	1.388 (0.172)	1.044 (0.305)
Stability test	Stable at 5% level	Stable at 5% level

Notes: * denotes 1% level of significance.

shows the result of the cointegration test for the Models 1 and 2. The F-statistics values for Models 1 and 2 are 9.438 and 9.507, respectively. Meanwhile, the values of T-statistics of Models 1 and 2 are −8.653 and −8.913, respectively. Since these values exceed the upper bound critical values at 1% significance levels, suggesting the rejection of the null hypothesis of no cointegration. Therefore, it is evident that there is a cointegrating association between ENVSU (CO₂ and EF) and its regressors in Egypt.

Furthermore, for the diagnostic analysis of each model (Table 4), we established that these models are free from issues such as misspecification of variables, non-normality, autocorrelation, and heteroscedasticity. Lastly, the residuals of each model were shown to be stable.

Table 5. ARDL estimators’ outcome.

	Model 1 (CO2 Emissions)		Model 2 (EF)
Variable	Coefficients	SE	Coefficients	SE
GDP	6.596 **	2.669	11.925 *	3.437
GDPSQ	−0.778 **	2.669	−1.659 *	0.499
DIGIT	−0.021	0.026	−0.317 *	0.054
HE	−0.583 *	0.205	−0.531 *	0.123
NGAS	−0.068 ***	0.038	−0.051	0.030
URB	0.090	1.169	−9.322	9.471
ΔHE	−0.380 *	0.106	−0.531 *	0.115
ΔDIGIT	-	-	−0.317 *	0.063
ΔURB			−9.322	6.097
ECT (−1)	−0.691 *	0.079	−0.739 *	0.082

Note: *, ** and *** denote 1%, 5% and 10% level of significance.

shows the result of the ARDL estimator for Models 1 and 2, respectively. The ECT(−1) is computed as −0.691 (Model 1) and −0.739 (Model 2) and is statistically significant at the 1% level, thereby validating the presence of long-term cointegration between ENVSU (CO₂ and EF) and its regressors in Egypt. It suggests that in any short-term disequilibrium, each model will converge to the long-run equilibrium at the rate of 69.1% and 73.9%. Furthermore, the results show that GDP exhibits a positive impact on ENVSU (CO₂ and EF) in Models 1 and 2. Particularly, for Model 1, the increase in GDP by 1% will increase CO₂ by 6.596%, while a percent rise in GDP will induce EF by 11.925%. Thus, GDP negatively influences ENVSU in Egypt. This agrees with prior studies such as [73] in Mexico, [74] in Italy, and [75] in Malaysia. Moreover, GDPSQ exerts a negative impact on ENVSU (CO₂ and EF) in Models 1 and 2. Specifically, increasing GDPSQ by 1% will lead to a decrease in CO₂ emissions by 0.778% (Model 1). Meanwhile, the surge in GDPSQ by 1% will reduce EF by 1.659%. This result established an inverted U-shaped association between economic growth and environmental quality in Egypt. The validity of EKC in Egypt stems from its shift from early-stage industrialization, which initially caused higher environmental degradation, to more recent investments in clean energy and environmental policies that have mitigated environmental degradation as the economy expanded. Our result is supported by prior studies, such as the works of [27,28], which confirmed the validity of the EKC in Türkiye. Additionally, the study of [38] also corroborated this result by revealing the validity of the EKC in Malaysia. Likewise, Shah and Ximei [39] also verified the presence of EKC in BRICST nations.

Furthermore, we observed that DIGIT does not significantly impact CO₂ emissions in Model 1. This disagrees with the work of [28], who suggested that industrial DIGIT negatively impacts CO₂ emissions in 40 nations. Likewise, Ullah et al. [14] observed the positive impact of DIGIT on CO₂ emissions in OECD nations. [30] also showed that DIGIT negatively impacts CO₂ emissions in European nations. However, in Model 2, there is a negative and significant linkage between DIGIT and EF in the long and short term. The surge in DIGIT by 1% will reduce EF by 0.137% in the long and short term. Thus, DIGIT negatively impacts EF in Egypt. This result coincides with the work of [26], who confirmed the positive role of DIGIT on ecological quality in G8 nations. Similar findings were also corroborated by prior studies, such as [29] in European nations and [27] in E7 nations, which revealed that DIGIT improves ecological sustainability. Adeshola et al. [15] emphasized that DIGIT contributes to ecological performance by optimizing resource usage and boosting energy efficiency across sectors with modern technologies such as AI and IoT. Additionally, DIGIT contributes to operational efficiency by incorporating renewable energy sources, propelling Egypt’s transition to a low-carbon economy and reducing its environmental impact. The implementation of digital platforms helps to increase environmental monitoring, which facilitates more effective regulatory enforcement and boosts sustainable practices.

Additionally, the results show that HE exerts a negative impact on ENVSU in the two models. Precisely, for Model 1, HE negatively impacts CO₂ emissions in the long and short term, wherein the rise in HE by 1% mitigates CO₂ emissions by 0.583% (long term) and 0.380% (short run). Our finding coincides with the work of [55], who observed that HE reduces CO₂ emissions in China. Ref. [56] also concluded that HE negatively affects CO₂ emissions in Turkey. Ref. [76] argued the negative role of HE on CO₂ emissions by providing a clean energy source that replaces fossil fuels, which are major contributors to CO₂ emissions during combustion. Furthermore, for Model 2, HE has an adverse impact on EF, whereby a percentage increase in HE will reduce EF by 0.531% (long and short term). Thus, HE induces ecological quality in Egypt. Other studies, such as [57] in Bangladesh and [58] in China, who observed that HE improves ecological quality. HE significantly reduces the EF by minimizing land degradation and habitat destruction associated with mining and drilling for coal, oil, and gas [77]. Moreover, the efficiency of HE plants in converting water flow into energy causes a reduction in resource usage compared to fossil fuel plants [78]. Cortés-Borda et al. [79] argued that the sustainable management of water resources in hydroelectric systems fosters biodiversity conservation, as it helps maintain aquatic ecosystems while mitigating the adverse effects of climate change.

It was observed that NGAS exhibits a decrease in CO₂ emissions, whereby the increase in NGAS by 1% will reduce CO₂ emissions by 0.068%. Thus, NGAS mitigates CO₂ emissions in Egypt. This result aligns with the work of [42], who indicated that NGAS reduces CO₂ emissions in China. Shah et al. [43] also argued for the decreasing effect of NGAS on CO₂ emissions in 13 selected nations. Meanwhile, prior studies such as [45,46] contradict our findings by suggesting that NGAS increases CO₂ emissions in India and France, respectively. NGA burns more cleanly than coal or oil, producing approximately 50% less CO₂ per unit of energy generated (IEA (2022) [80]). This cleaner combustion process helps Egypt transition toward lower-carbon energy production while still meeting growing energy demands. Additionally, the efficiency of NGAS plants, particularly with combined-cycle technology, further reduces CO₂ emissions by converting more of the fuel’s energy into electricity (EIA (2021) [81]). For Model 2, NGAS exerts a negative and insignificant impact on EF in Egypt. This result suggests that the NGAS does not have a significant impact EF. Despite NGAS being a cleaner alternative compared to coal or oil; its usage may not significantly mitigate environmental damage. Moreover, NGAS contributes to energy production, its impact alone is insufficient to significantly alter EF without comprehensive systemic reforms in environmental and energy laws.

Lastly, the results show that URB exerts an insignificant impact on ENVSU in both models. Thus, this result suggests that the current URB in Egypt does not significantly influence the country’s environmental quality. This finding indicates that urban expansion, in its present form, may not be a major driver of environmental degradation, possibly due to effective urban planning or lower industrial emissions within urban areas. It also implies that other factors, such as energy production methods, agricultural practices, or transportation systems, might have a more substantial impact on environmental quality, overshadowing the direct effects of URB. These findings do not align with prior studies such as the work of [40], who revealed that URB contributes to ecological sustainability. The findings of other studies such as the work of [47], who conducted studies in ECOWAS nations, suggested that URB negatively impacts EF. Likewise, the study of [22] in Finland and [52] in China also established that URB negatively impacts CO₂ emissions. Ref. [48] further argued that URB positively impacts EF in Russia.

Table 6. Causality outcome.

	Long Term		Medium Term		Short Term
	T-Stat	p-Value	T-Stat	p-Value	T-Stat	p-Value
GDP → CO2	6.278 **	0.043	3.590	0.166	0.947	0.623
CO2 → GDP	4.707 ***	0.095	6.024 **	0.049	9.505 *	0.008
EF → GDP	7.024 **	0.029	11.028 *	0.002	10.375	0.005
GDP → EF	8.560 **	0.013	0.805	0.668	0.219	0.896
NGAS → CO2	2.648	0.266	5.173 ***	0.075	2.404	0.300
CO2 → NGAS	10.992	0.004	8.399 **	0.015	8.840 **	0.012
NGAS → EF	1.075	0.584	8.093 **	0.017	8.062 **	0.018
EF → NGAS	6.487 **	0.039	0.596	0.742	0.517	0.772
CO2 → HE	8.244 *	0.016	3.527	0.171	2.853	0.240
HE → CO2	11.359 *	0.003	5.005 ***	0.082	2.336	0.311
HE → EF	9.853 *	0.007	8.863 **	0.012	2.734	0.255
EF → HE	5.126 ***	0.077	5.471 ***	0.065	6.812	0.033
URB → CO2	5.751 ***	0.056	7.151 **	0.028	9.067 **	0.011
CO2 → URB	5.885 ***	0.052	3.076	0.215	3.104	0.212
URB → EF	0.521	0.770	5.444 ***	0.065	5.212 ***	0.073
EF → URB	3.393	0.183	8.651 **	0.013	4.190	0.123
CO2 → DIGIT	8.896 **	0.012	5.273 ***	0.072	4.657 ***	0.097
DIGIT → CO2	7.298 **	0.026	2.785	0.248	1.909	0.384
DIGIT → EF	4.892 ***	0.086	5.110 ***	0.077	1.063	0.587
EF → DIGIT	5.886 ***	0.053	6.418 **	0.040	5.554 ***	0.062

*, **, and *** portray significance levels of 0.01, 0.05, and 0.1.

shows the result of the frequency domain causality approach. We observed that GDP can predict the values of both CO₂ and EF in the long term. Likewise, EF can also predict the value of GDP at all periods. Thus, there is a two-way causal linkage between environmental sustainability and GDP in Egypt. Additionally, a bidirectional causal association between NGAS and environmental sustainability in Egypt is observable. Precisely, NGAS can predict the values of CO₂ emissions in the medium term, while CO₂ emissions can predict the value of NGAS in the medium and short term. Moreover, NGAS is a predictor of the values of EF in the medium and short term, while EF can predict the value of NGAS in the long term. HE granger causes CO₂ and EF in the long and medium term. Additionally, CO₂ emissions can predict HE in the long term, whereas EF is also a predictor of HE in the long and medium term. Thus, HE and ENVSU both granger causes each other in Egypt. A causal connection between URB and CO₂ was observed across all periods, while CO₂ granger causes URB in the long term. Likewise, URB is a predictor of EF in the medium and short term, while EF can also predict the value of URB in the medium term. Thus, there is a bidirectional causality connection between URB and ENVSU in Egypt. Lastly, DIGIT can predict the values of CO₂ emissions in the long term, while DIGIT can forecast CO₂ emissions across different periods. Also, DIGIT granger causes EF in the long and medium term, while EF is a predictor of EF across different periods. Hence, it is evident that there is a bidirectional causal association between DIGIT and environmental sustainability in Egypt.

5. Conclusions and Policy Remarks

This research utilized the ARDL estimator to probe how hydroelectricity, digitalization, natural gas, and urbanization influence CO₂ emissions and EF within the EKC hypothesis in Egypt. Upon conducting a series of unit root tests such as ADF and PP unit root tests, this study further conducted the bound testing method and revealed that there is a long-term cointegrating association between environmental sustainability (CO₂ emissions and EF) and its regressors. The findings of the ARDL estimators showed that GDP positively impacts environmental sustainability (CO₂ emissions and EF), while GDPSQ negatively influences environmental sustainability (CO₂ emissions and EF), indicating that there is an inverted U-shaped association between GDP and environmental sustainability (CO₂ emissions and EF) in Egypt. Additionally, URB does not affect environmental sustainability (CO₂ emissions and EF), while HE negatively impacts environmental sustainability (CO₂ emissions and EF) in Egypt. DIGIT does not significantly impact CO₂ emissions, while DIGIT decreases EF. Meanwhile, NGAS mitigates CO₂ emissions in Egypt, while NGAS does not significantly impact EF. Furthermore, the frequency domain causality approach showed that there is a two-way causality pathway between environmental sustainability (CO₂ emissions and EF) and its regressors.

Policy Remarks

To effectively balance economic growth with environmental sustainability in Egypt, policymakers must adopt a nuanced and sophisticated strategy that leverages the dynamic relationship between GDP and environmental outcomes. The inverted U-shaped association between GDP and sustainability suggests that Egypt is at a crucial stage where further economic development can either exacerbate environmental degradation or shift toward greener growth. Consequently, the government should prioritize decoupling economic growth from CO₂ and EF by investing in green infrastructure, sustainable industries, and circular economy models. Introducing fiscal incentives, such as tax breaks and subsidies for clean energy technologies, eco-friendly manufacturing, and sustainable agriculture, will accelerate the transition toward a low-carbon economy, fostering a more sustainable trajectory for GDP growth.

Moreover, Egypt must focus on transitioning from resource-intensive growth to knowledge-based, innovation-driven growth. The expansion of digital infrastructure, as indicated by the positive role of DIGIT in reducing EF, should be coupled with policies that promote green technological innovation. Research and development in clean technologies, including renewable energy, carbon capture and storage, and energy-efficient industrial processes, should be prioritized through public–private partnerships. By creating a regulatory environment that encourages sustainable entrepreneurship and green business practices, Egypt can position itself as a leader in eco-innovation within the MENA region. Additionally, policies fostering education and skill development in green sectors are essential to equip the workforce with the competencies needed for a sustainable economy.

Simultaneously, Egypt’s policy framework should address the negative environmental impacts associated with HE and ensure that NGAS continues to play a strategic role in reducing CO₂ emissions. While natural gas offers a cleaner alternative in the short to medium term, Egypt must prepare for long-term economic growth by gradually diversifying its energy portfolio with a higher share of renewables, such as solar and wind. Investments in green finance mechanisms, like carbon pricing and green bonds, can provide the capital needed to scale up clean energy initiatives and enhance the financial sustainability of environmentally responsible projects. Ultimately, aligning economic growth strategies with sustainability targets will enable Egypt to achieve long-term prosperity while mitigating environmental risks, fulfilling international climate commitments, and improving the overall quality of life for its population.