Research on Geopolitical Risk

In recent years, the world economic landscape and geopolitical situation have stood tremendous challenge [8]. According to Caldara and Iacoviello (2022), geopolitical risk is defined as the risk associated with war, terrorism, and tensions affecting relations between states. It is separated into two categories: risks connected to recent events and risks associated with recent events’ escalation. Military wars, ongoing armed conflicts, and political turmoil have increased global geopolitical risks, which are destructive to both national and international economic growth and energy security. In general, areas with abundant oil and other fossil fuels, tend to have higher geopolitical risk, such as the Middle East, North Africa, and Russia. Therefore, geopolitical risk is going to have a negative influence on the growth of the world’s energy markets and systems.

The conflict between Russia and Ukraine is another example of how geopolitical issues have had an adverse effect on energy production and consumption. Apart from higher energy prices, the conflict has adversely affected the stock market and energy consumption and production. The Granger causality model was developed by Fernandois and Medel (2020) to investigate the influence of geopolitical conflicts and unforeseen alterations in the price of oil [9]. According to Li et al. (2020), geopolitical risk has a favorable impact on oil prices for the West Texas Intermediate (WTI) and Brent markets but has the reverse effect on the Dubai and Nigerian markets [10]. This illustrates that the results of geopolitical risk vary across diverse areas. There are various ways in which geopolitical factors might have a huge effect on the oil markets.

In addition, the negative influence of geopolitical risk on energy security and renewable energy development is the focus of existing research. Khan et al. (2020) explained that economic and political crises have worsened competition in energy markets [11]. This competition increases the instability of the energy supply. Insecurity and geopolitical risk in resource-rich countries caused oil prices to rise in the short term and oil production to fall sharply. Gong et al. (2022) studied a system to estimate the safety of China’s oil fields. Then, explored how geopolitical risk affected security [12]. They pointed out that geopolitical risk endangers the sources of China’s oil imports and have a negative impact on the security of China’s oil resources. What’s more, from 1985 to 2018, the generation of energy from renewable sources was significantly impacted by worldwide geopolitical risk. Floros et al. further confirmed that the relationship between international politics and energy is considered an important element for economic development from multiple perspectives such as economic growth, climate change, and energy security [13]. In further research, geopolitical risk boosted the profitability of clean energy pricing in a brief period, according to research by Sarker et al. using data collected in time series from 2001 to 2021. On the other hand, over time, the impact is detrimental [14].

Research on Energy Resilience

As a significant precondition for driving economic growth, stabilizing employment, and promoting sustainable development, energy is already a key factor in promoting industrialization and urbanization [15]. In recent years, resorting to renewable energy gradually attracted the attention of countries around the world and has become a mainstream trend, partially replacing traditional fossil fuels. Considering the instability and unavailability of renewable energy, and the random and ever-increasing characteristics of energy demands, it is necessary to assess the intrinsic features of district energy systems, especially for energy resilience during periods of exposure to external risks and shocks.

Recently, the study on the issue of energy resilience has gradually increased, yet there isn’t an explicit description or way to quantify energy resilience in the research currently available. Energy sustainability, energy vulnerability, and resilience have not been properly distinguished from one another. In the context of the gradual depletion of natural resources and the rapid development of renewable energy, Niu et al. deeply explored the causes and countermeasures of energy poverty, emphasizing the strategic importance of energy resources and the need to enhance energy resilience [16]. In addition, Feng et al. analyzed the influencing factors of energy resilience from the perspective of energy security, and put forward suggestions for optimizing energy issues and ensuring energy security and sustainable development [17]. According to Gatto and Drago (2020), energy resilience in our research might be defined as a system’s capacity to react to and recover from external shocks and risk shocks (such as economic, social, environmental, and public health events). In order to understand the characteristics of energy resilience, different scholars use various methods to measure the energy resilience. For instance, Gupta et al. (2019) indicated that the energy resilience of a deprived neighborhood in Oxford by deploying solar PV systems and smart batteries for 82 residential communities [18]. Shen et al. studied the impact of industrial agglomeration on carbon emissions and proposed that changing the way energy is produced and used and investing in renewable energy are important ways to reduce carbon emissions and improve energy resilience [19]. To maintain sustainable development, Guang et al. (2019) used China’s multi-regional input-output table in 2012 to propose an input-output linear programming model, which can analyze the energy elasticity of multi-regional economies. The authors found that technical efficiency in energy-intensive industries has a significant impact on energy resilience [20]. On that basis, Gatto and Drago (2020) developed an interval-based composite energy resilience indicator using the three parts and 27 indicators of the Sustainable Energy for All initiative offered by the World Bank (2019). Moreover, Feng et al. explored the impact of low-carbon energy transition on energy intensity, which can effectively reduce energy intensity by improving the use of renewable energy, reduce dependence on traditional energy, and enhance the resilience of the energy system [21].

In short, the papers mentioned above demonstrate that researchers have evaluated energy resilience from several professional angles. However, a thorough and efficient analysis of energy resilience from the standpoint of influencing factors is usually neglected. Meanwhile, geopolitical risk and other factors affecting energy resilience have not attracted extensive focus.

Research Between Geopolitical Risk and Energy

At present, there are almost no papers analyzing the influence of geopolitical risk on energy resilience, and some existing studies mainly focus on the relationship between geopolitical risk and energy safety or the connection between geopolitical risk and renewable energy development. Regarding these studies, for instance, Su et al. analyzed the impact of geopolitical risk on oil supply security, pointing out that oil supply security depends on geopolitical risk [22]. In this context, oil-rich regions are highly vulnerable to geopolitical risk, as reflected in rising energy security indices. The conclusion shows that geopolitical risk contributes to oil supply security. In a further study, Vakulchuk et al. (2020) investigate that geopolitical risk aggravates security risks and international relations [23]. Similar to this, Su et al. believed that geopolitical risk and energy had a mutually reinforcing connection [24]. Zhang et al. studied the global causal relationship between renewable energy consumption, financial development, and public health in the context of changing geopolitical risks, emphasizing the importance of renewable energy, which provides an important reference for accelerating energy transition and enhancing energy resilience [25]. Additionally, Li et al. (2022) concluded that geopolitical risk played a critical role in the insecurity of the energy supply [26]. Feng et al. studied the negative impact of geopolitical risks on environmental changes and concluded that geopolitical risks pose a threat to energy security and supply stability in countries rich in natural resources, which is a great challenge [27].

Furthermore, Sweidan et al. (2021) studied the effect of geopolitical risk on renewable energy development in developing economies with net crude oil imports. The study found the influence of geopolitical risk on the deployment of renewable energy in America between 1973 and 2020. The results showed that geopolitical risk promoted the deployment of renewable energy in America. Sweidan checked the effect of geopolitical threats on renewable energy utilization and uncovered that geopolitical threats would promote the utilization of renewable energy. Shen et al. studied the impact of environmental regulation on energy in the context of geopolitical risks and carbon neutrality goals and concluded that informal environmental regulation can reduce carbon emissions by promoting industrial structural transformation and renewable energy substitution, which provides an effective reference for enhancing energy resilience [28]. Zhao et al. (2023) investigated how geopolitical risk affected the requirement for renewable energy in 20 OECD member nations from 1970 to 2019. Geopolitical risk is proven to decrease consumption of renewable energy and jeopardize the strategies for reducing climate change. Additionally, policymakers in OECD nations should consider regional geopolitical harmony [29].

In brief, a number of studies in the literature concern the influence of geopolitical risk on energy security and renewable energy. These papers mainly start from the aspects of energy supply security or renewable energy demand, so as to deeply explore and analyze the influence of geopolitical risk on the energy system. In this study, we through the analysis from 2000 to 2019, 20 countries geopolitical risk panel data set for the impact of higher energy resilience, contributing to the academia. To our knowledge, this is the first paper in the literature that examines the connection between geopolitical risk and energy from the perspective of energy resilience.

Literature Gaps

Undeniably, the extant literature on geopolitical risk and energy resilience is sufficient and substantial. Nevertheless, the influence of geopolitical risk on energy resilience has received only a small amount of research attention to date. Energy resilience has been defined and evaluated in numerous prior papers, but the academic field has not yet created an official, common standard. The establishment of a fair and efficient index system for evaluating energy resilience is necessary. Additionally, to our knowledge, no research has looked at how geopolitical concerns affect energy resilience. Analyzing the diverse impacts of geopolitical risk on energy resilience is essential to ascribe to the intrinsic differences in economic and natural conditions across countries. Furthermore, it is essential, from the perspective of building an independent and complete energy system, to explore the unique mechanism of geopolitical threats on energy resilience from the perspective of scale effect, structural effect, and technological effect. However, there haven’t been any comprehensive analyses on the previously mentioned subjects by researchers.

ERI Measurement

In order to effectively assess energy resilience and further investigate the influence of geopolitical risk on it, the energy resilience calculation index is built into this study. This study further estimates the energy resilience index (ERI) statistics for every country from 2000 to 2019 using the energy vulnerability index (EVI) data. We observe that, in terms of numerical values, the ERI and the EVI show a trend of rising and falling.

The IEM technique is employed in this study, which is based on Dong et al. (2021), to determine an integrated energy vulnerability index by giving each sub-indicator a distinct weight [30]. Then, the energy resilience is further determined. The steps in the computation are as follows.

To get proper weights, each sub-indicator must first undergo standardization in order to remove the impacts brought on by unit discrepancies [31]. The seven positive and nine negative proofs of the indicator system are normalized using Eq. (1) and (2):

Second, on the basis of the standardized data of each sub-indicator, the information entropy is calculated as follows:

Third, the weight of each sub-indicator, i.e., the degree of contribution to the measurement result, is calculated as follows:

Lastly, a composite energy vulnerability index is created by weighting and summing each dimensionless sub-indicator. Then, the energy vulnerability composite index is subtracted from 1 to obtain the energy resilience composite index:

In conclusion, the energy resilience index (ERI) has a range of 0 to 1. Energy resilience increases with increasing ERI, and vice versa.

Econometric Model

From the angle of the Porter hypothesis, we analyzed the influence of GPRI on the ERI by adopting a fixed effect (FE) regression estimation model. The model can reduce estimation bias from omitted variables. The specific econometric model was expressed as below:

Whereas, ERI_it and GPRI_it are the explained and core explanatory variables of i country at t period, respectively. X_it are the control variables. μ_i and v_t denote the individual and time effect. ε_it is the random error.

Variable Selection

(1) Dependent Variable (ln ERI_it). The dependent variable is the energy resilience of countries, which is calculated based on the data of the EVI. Following the studies of Gatto and Busato (2020) and Genave et al. (2020), we also consider the impact of energy mix diversity, mainly measured by the Shannon-Wiener index. At the same time, we construct an energy security indicator including energy import, fuel export, and energy structure diversity as the key sub-indicator to measure the energy vulnerability index (EVI), which can be further calculated to obtain the energy resilience index (ERI).

(2) Core independent variable (lnGPRI_it). The independent variable comes from British Petroleum (2022), and geopolitical risk is the main explanatory factor. Caldara and Iacoviello (2022) provided the geopolitical risk index (GPRI) statistics on their website. In particular, the GPRI’s coefficient is predicted to be positive, indicating that lowering geopolitical risk also lowers energy resilience. In addition, the decline of energy resilience means that the energy intensity is reduced, which is the embodiment of the beginning to change the extensive development, so as to further promote the optimization and upgrading of energy structure.

(3) Control variables. To mitigate the influence of the omitted variables on the estimated results, this paper selected the following indicators as control variables. Carbon emissions per capita (lnPCO2_it) was measured by carbon emissions per capita after taking logarithm. Numerous studies have investigated the connection between carbon emissions and energy, considering that the significance of decreasing carbon emissions to combat global climate change encourages nations to switch from fossil fuel energy to RE. The carbon emissions from consumption of energy were used in this study as a proxy for carbon emissions. The per capita GDP was converted to log form (lnPGDP_it ) to determine the level of affluence. Geopolitical risks are political and economic trends that are potentially destructive to human well-being. In addition to violent conflicts, the most important manifestations are global or regional economic and financial turbulence and natural energy risks caused by non-human environmental changes. From an economic perspective, the intensification of geopolitical risks will directly cause economic problems and lead to changes in GDP per capita. In terms of energy, generally speaking, as the level of economic development increases, the energy resilience of a region will increase. However, the economic development experience of developed countries shows that with economic development, the trajectory of the energy resilience coefficient changes in an inverted N-shaped shape. Hence, it is theoretically uncertain what relationship energy resilience and economic growth should maintain in different periods of economic development. The value of natural resources (such as petroleum, natural gas, coal, minerals, and forests) to economic output is measured by natural resource rents (NATRES_it) as a portion of gross domestic product (GDP). Natural resource rents are typically derived from oil and gas rents. According to Sweidan and Elbargathi (2022), globalization and natural resource (oil) rents are important controls for examining the impact of geopolitical risks on economic development. Therefore, we choose natural resource rents as one of the control variables for our empirical analysis. Data on GDP per capita and natural resource rents are taken from the World Bank’s World Development Indicators dataset (2023). Following Dong et al., we use the energy vulnerability index to proxy energy system level (lnEVI_it) and express it in logarithmic form. The EVI has a range of 0 to 1. Energy system vulnerability increases with increasing EVI values and vice versa. Notably, the energy resilience index and the energy vulnerability index show a trend of rising and falling, correspondingly. Meanwhile, we also choose the log form of total primary energy supply (ln TPES_it) as a control variable. Fossil fuels account for 82.28% of the total energy supply globally in 2021. The continued dominance of fossil fuels in the global energy mix means that an increase in the overall amount of energy available will ultimately result in higher energy use and higher CO2 emissions. In addition, we explore how energy resilience, which is determined by the output of the economy (i.e., GDP) corresponding to a unit of energy consumption, is affected by energy efficiency (ln EE_it). According to Dong et al. (2020), fossil fuels tend to explain a sizable portion of the world’s energy consumption and are a significant contributor to carbon emissions [32]. Therefore, increasing energy efficiency is crucial for reducing carbon emissions globally and affecting energy resilience. Energy efficiency improvements are successful at reducing CO₂ emissions, according to research by Akram et al. (2020), and as a result, have some indirect influence on energy resilience [33]. Additionally, Energy efficiency has been also utilized as a control variable in papers by several researchers [34]. The specific variable construction and sources are shown in Table 1.

Data Source

Based on the data continuity and availability, we employ balanced panel data for 20 countries from 2000 to 2019 for the empirical regression, which is the maximum sample size that can be selected due to data constraints. These countries contain almost all the major countries distributed in all continents of the world and can effectively represent the whole sample. Missing data is the only reason for dropping Spain, Japan, and other countries from the sample. The variables mentioned above were collected from the International Statistics Yearbook. Table 2 displays the descriptive statistics of the explained and explanatory variables. The table contains the fundamental facts, such as observed quantity, minimum, mean, maximum, and standard deviation. This information illustrates the concentration and distribution of data.

Differential-Generalized Method of Moments (Diff-GMM)

The following supports the selection of Diff-GMM as the benchmark model: First, in contrast to the static model, the dynamic model considers the impact of the lagged energy resilience features. The dynamic model may resolve the endogeneity issue because of two-way causality. However, the inclusion of the dependent variable’s lag term results in the development of a fresh endogeneity issue. Arellano and Bond (1991) suggested utilizing a Diff-GMM strategy to overcome this endogeneity problem, in which all potential lagged variables are used as instrumental factors. In their Sys-GMM proposal from 1998, the research of Arellano and Bover (1995) was extended by Blundell and Bond by analyzing distinctions and level equations as a set of equations. Diff-GMM has a greater estimation efficiency than Sys-GMM and is better able to deal with a possible poor instrumental variable difficulty.

The Arellano-Bond (A-B) estimation and the Hansen test must also be used on the assessment findings in order to confirm their accuracy. The differences of the random disturbance terms must have first-order autocorrelation and no second-order autocorrelation in order for the A-B test to be valid, meaning that the p-values for AR (1) and AR (2) should, respectively, be less than 0.1 and larger than 0.1. The Hansen test is used to test for over-identification of the instrumental variables, and a p-value of more than 0.1 ensures that all the instrumental variables are valid.

Assessments from Table 6 indicate a substantial positive relationship between GPRI and ERI, suggesting that lowering the geopolitical risk may simultaneously decrease energy resilience. This new information supports the demand for actions to increase the energy system’s capacity to handle geopolitical risk. It turns out that each 1% drop in GPRI can result in a 0.17% drop in energy resilience, which is in line with our estimate. A possible reason is that the decline in geopolitical risk discourages energy resilience by reducing non-essential energy rivalry and raising energy efficiency [37]. Nevertheless, Dong et al. (2021) reached the opposite finding in a related investigation. Strong energy resilience may be able to ensure energy supply in time to withstand extrinsic risks, which could explain why the geopolitical risk has decreased. However, with the growth of science and technology, energy efficiency is gradually improved, and the transformation process of energy structure is accelerated, which enhances the ability to deal with geopolitical risk and decreases energy resilience in the long run, achieving a “win-win” condition.

Alternative Dependent Variables

Although researchers usually apply energy resilience to analyze the energy resilience of an area, many researchers consider renewable energy consumption (REC), measured in exajoules. They believe that renewable energy consumption is very appropriate to explore the energy resilience of a region because it considers the factors of energy demand. Therefore, REC is used as a new dependent variable in this section’s regression analysis, and the results are shown in column (1) of Table 7. The predicted GBP coefficient is strongly positive, which is in line with the results of the benchmark regression.

Effect of Lagging One Period

Geopolitical risks typically take time to materialize before creating an effect on social production and economic activities. The pertinent findings are therefore in column (2) of Table 7 when using ERI(t+1) as the dependent variable to reevaluate the impact of geopolitical concerns. The positive predicted GPRI coefficients suggest that geopolitical risk is advantageous for boosting energy resilience. The aforementioned findings demonstrate the main conclusion is robust.

Adding Control Variables

By including control variables, this section reduces the bias in the results brought on by the omission of certain variables. We discover the influence of the evolution of urbanization (URB) as measured by the urban population’s share of the total population. Meanwhile, we assess the influence of trade openness (TRA) on energy resilience, as measured by the ratio of total imports and exports to GDP. The relevant data are shown in column (3) of Table 7 after considering the two controls for URB and TRA. The coefficient of geopolitical risk is positive, which is in accordance with the main result of the study.

Winsorizing the Variables

Outliers can cause alters in the main trend, which could lead to erroneous conclusions. Hence, this section uses the winsorizing 1% quantile method for data smoothing in accordance with convention and the number of samples. The pertinent regression results are shown in column (4) of Table 7. The underlying conclusion is more robust because the GPRI coefficients are still positive.

Elimination of Time Periods

To exclude the impact of the 2008-2009 financial crisis on the sample data, we adopt the elimination method of time period, so as to better analyze the impact of geopolitical risks on energy resilience. The relevant regression results are shown in Column (5) of Table 7. Since the coefficient of GPRI is still positive, the basic conclusion is consistent with the previous one.

Model Construction

Nevertheless, the precise internal influence mechanism is unknown, which attracts our curiosity in this field of study. The study hypothesizes that geopolitical risks have a direct positive effect on energy resilience and that the total primary energy supply (TPES), the proportion of renewable energy consumption in the overall energy supply (REC), and energy efficiency (EE) mediate this relationship through geopolitical risks. We employ the mediation analysis using a stepwise regression approach to examine this hypothesis and utilize the Sobel test to assess the significance of the mediating effect. As a result, we create the theoretical model below to investigate the impact processes, referencing the research of Baron and Kenny (1986).

In the above formula, η₀ and γ₀ represent the constant terms; η₁ − η₇ and γ₁ − γ₈ stand for the estimated coefficients respectively; ε_it stands for the random disturbance term; and M_it for the mediating variables.

According to earlier research, the shape effect, structural effect, and technological effect can all be used to explain energy resilience. Thus, we empirically investigate how the three effects affect the link between geopolitical risk and energy resilience. In order to depict the scale effect, we first select the total primary energy supply (TPES) as a mediating variable. According to Roupas et al. (2011), 82.28% of the world’s primary energy will come from fossil fuels in 2021. Since fossil fuels still account for the majority of the world’s energy mix, an increase in the total amount of available energy will inevitably lead to higher energy consumption and thus higher carbon dioxide emissions. Secondly, in order to capture the structural influence, we select the proportion of renewable energy consumption in the overall energy supply (REC) as a mediating variable. According to Dong et al. (2020), renewable energy plays a crucial part in the mitigation of the greenhouse impact. The share of primary energy coming from renewable sources has increased significantly in recent years, from 7.82% in 2000 to 13.47% in 2021. Third, for the sake of capturing the technological effect, we select energy efficiency (EE) as a mediating variable. Global energy intensity has dramatically lowered as a result of technological advancements, falling from 1.84 kWh/USD in 2000 to 1.42 kWh/USD in 2018. Energy resilience is impacted even more by increases in energy efficiency since they not only lower the expense of power generation but also motivate enterprises to upgrade their present technology through competition [39]. We decided to apply energy efficiency as a stand-in for the technological effect because Wang et al. ‘s (2022) found that it can be a reliable predictor of technological advancement. The UN statistics and Worldwide Development Indicators databases provide the pertinent information for the mediating variables.

Additionally, if γ₂ in Eq. (10) is obvious, it points out that GPRI directly and significantly influences energy resilience; if γ₃ in Eq. (10) and η₂ in Eq. (9) are both significant, it shows that geopolitical risk also indirectly and significantly influences energy resilience.

Results and Discussions

The results of mediating effects are shown in Table 10. Columns (1) and (2) represent the scale effect. The findings indicate that GPRI and TPES have a positive association and that a 1% increase in TPES causes a 0.25% increase in energy resilience. This suggests that GPRI can indirectly enhance energy resilience by increasing the entire primary energy supply. Although it is an inevitable trend for new energy sources, mainly wind, electricity, and light energy, to replace fossil energy, the dependence of the energy market on high-carbon fossil energy is difficult to be shaken in the short term. The energy consumption of various countries mainly depends on fossil energy, mainly coal, automobiles, and steel, and the continuous burning of fossil energy promotes the expansion of the energy scale [40]. However, due to the stationarity of fossil energy supply, the dependence of the energy market on fossil energy acts as a stabilizer, which ensures that the energy system has stable and sufficient energy reserves, thus indirectly enhancing energy resilience.

The structural effect is outlined in columns (3) and (4). We discover a substantial positive relationship between GPRI and REC, and that for every 1% rise in REC, the energy resilience decreases by 0.057%. This suggests that GPRI indirectly reduces energy resilience under an increase in renewable energy use. In the face of geopolitical risk and other external shocks, some countries tend to use new energy to replace traditional fossil energy. However, there is great uncertainty in the acquisition of new energy, mainly wind and light energy, which makes energy supply and demand also fluctuate to some extent. Therefore, the process of energy structure reconstruction is not smooth. Geopolitical risk impacts the transformation of energy structure, destroys the process of energy system reconstruction, and reduces energy resilience. On the other hand, under the general trend of energy structure transformation, countries are optimizing and upgrading their energy structure by replacing old energy such as coal and fossil energy with new energy [41]. However, at the same time, the labor and cost of energy structure adjustment are also huge, and in the face of the impact of external geopolitical risk, there is great instability in energy supply and demand, which makes it inevitable to reduce energy resilience. Hence, we demonstrate that a key mediating factor in the relationship between GPRI and energy resilience is renewable energy.

Columns (5) and (6) represent the technical effect. Based on the analysis of the relationship between GPRI and EE, it is concluded that there is a positive correlation between geopolitical risk and energy efficiency, and every 1% increase in energy efficiency will decrease energy resilience by 0.141%. This suggests that by increasing energy use efficiency, GPRI can indirectly reduce energy resiliency. This discovery strengthens the argument that energy efficiency plays a crucial role in ensuring energy resilience, which has been widely acknowledged by academics. Many governments have recognized that increasing energy efficiency is a good energy policy in order to alleviate the effects of energy issues. For instance, China has made enhancing energy efficiency a top priority in order to achieve its goals of carbon peaking and carbon neutrality; the European Commission has also made enhancing energy efficiency a top priority in order to achieve its goals of the European Green Deal. However, in terms of economic costs, improving energy efficiency will not happen overnight. In the process of improving energy efficiency, countries are constantly trying to replace old energy with new energy and promote the adjustment of energy structure. Due to the long investment cycle, energy supply and demand are unstable, which indirectly impacts the energy system and reduces energy resilience. Subsequently, we conduct the Sobel test to examine the significance of the mediating effect. The results of the Sobel test show that the Sobel statistics of the scale effect, the structure effect, and the technology effect all exceed the critical value of 1.96 (α = 0.05), which indicates that the mediating effect is statistically significant.

Last but not least, when we compare the GPRI coefficients in columns (2), (4), and (6) to those in the benchmark regression, the paper discovers that not only are they all less significant than the coefficients in the benchmark regression, but they are also all smaller. This indicates that scale effect, structure effect, and technology effect all play a mediating role to a certain extent, and geopolitical risk has a significant impact on energy resilience but does not have a significant impact on the mediating variable. In other words, the scale effect, the structure effect, and the technology effect all indirectly affect energy resilience, while the significance is weak and different [42]. The main reason is that in the process of energy structure reconstruction, the dependence of the energy market on fossil energy makes the scale effect strong and plays a role in enhancing energy resilience. Nevertheless, due to the uncertain characteristics of new energy, such as the supply of light energy, wind energy, and electric energy will be disturbed by many factors, and the supply will be unstable [43]. Therefore, in the process of promoting energy structure adjustment, both the structural effect and the technical effect play a negative role, which hinders the promoting effect of geopolitical risk on improving energy resilience.

In summary, in the context of increasing geopolitical risk, increasing total primary energy supply can indirectly enhance energy resilience, which is mainly attributed to the dependence of the energy market on high-carbon fossil energy. Conversely, increasing the use of renewable energy and improving energy efficiency indirectly reduce energy resilience, and the possible reason is the uncertainty of new energy supply and high investment cost during the restructuring of the energy structure. It reflects the challenges encountered in the process of adjusting the energy system, which aligns with empirical evidence and underscores the scientific and rational nature of this study.