Transition Energy and Technical Efficiency of Energy Companies: DEA and Panel Evidence from Renewable and Traditional Energy Companies in Europe and North America

Abstract

This study examines the technical and operational efficiency of publicly listed energy companies operating in Europe, the United States, and Canada during the 2017–2024 energy transition period. The sample includes both traditional electricity utilities and renewable energy producers. Technical efficiency was estimated using output-oriented Data Envelopment Analysis (DEA), specifically the Charnes–Cooper–Rhodes (CCR) and Banker–Charnes–Cooper (BCC) models. Panel-data models were subsequently applied to identify the financial, organisational, regional, and environmental, social, and governance (ESG) factors associated with firm-level efficiency. The results indicate a moderate average level of technical efficiency, with a substantial share of inefficiency attributable to an inappropriate operating scale. Contrary to the initial hypothesis, renewable energy companies were, on average, less technically efficient than traditional utilities, despite achieving higher ESG and environmental scores. European companies exhibited higher efficiency than firms located in the United States and Canada, suggesting that long-term exposure to climate-policy and regulatory pressures may encourage more effective resource use. The panel-model results did not provide robust evidence that ESG performance directly improves technical efficiency. By contrast, profitability, leverage, and firm size were significantly associated with efficiency outcomes. These findings show that the energy transition depends on more than the expansion of renewable energy capacity. Effective resource allocation, financial resilience, organisational adjustment, and an appropriate operating scale are equally important. The study provides relevant implications for corporate managers, investors, and policymakers involved in energy-sector transformation.

1. Introduction

The energy sector is currently experiencing very dynamic changes related to the energy transition, climate policy, and the growing importance of renewable energy sources. Today, energy companies must supply energy, maintain financial stability, and adapt to increasingly demanding environmental regulations and sustainability expectations. In practice, this means simultaneously achieving economic, environmental, and organisational goals, which poses a significant challenge for many companies [1].

In recent years, the question of whether companies that implement an energy transition are truly more efficient than traditional energy companies has been increasingly raised, as well as whether a high level of environmental commitment translates into better economic performance. To date, research has not provided a clear answer. On the one hand, it indicates that investments in modern technologies and ESG can improve corporate efficiency, but, on the other hand, the energy transition is associated with high investment costs, which can reduce the profitability and operational efficiency of companies in the short term [2].

The objective of this study is to assess the technical and operational efficiency of energy companies operating in Europe, the United States, and Canada during the ongoing energy transition. Particular attention was paid to comparing renewable energy companies and traditional energy companies, analysing the importance of ESG, and evaluating the impact of regional and organisational factors on the efficiency of energy sector companies.

The study will analyse the following:

• The level of firm-level technical-operational efficiency of energy companies;
• Differences between renewable energy companies and traditional energy companies;
• The influence of the region of operation (Europe vs. the USA and Canada);
• The importance of the scale of the companies’ operations;
• The influence of ESG indicators and environmental factors on technical efficiency;
• Determinants of company efficiency in panel models.

The analysis was carried out for energy companies between 2017 and 2024. The study used a combination of DEA methods (CCR and panel BCC) and fixed-effects (FE) models. This approach makes it possible to assess company efficiency and identify its determinants. DEA models enabled the assessment of companies’ ability to generate operating results with a given level of financial and asset resources, while panel models allowed for the analysis of the impact of ESG, region of operation, and company type on the level of technical efficiency. The following research hypotheses were tested in the study:

• H1: Renewable energy companies achieve higher technical efficiency than traditional energy companies;
• H2: European companies achieve higher technical efficiency than companies from the US and Canada;
• H3: Higher ESG indicators positively impact the firm-level technical-operational efficiency of energy companies;
• H4: A significant portion of energy company inefficiency results from suboptimal scale of operations;
• H5: The energy transition affects company efficiency through both environmental and organisational-financial factors.

The main contributions of this study are as follows.

• It combines DEA efficiency analysis with panel regression models within a single empirical framework.
• It compares renewable and traditional energy companies operating under different regulatory environments in Europe and North America.
• It assesses technical operational efficiency during the energy transition period (2017–2024), including years characterised by the energy crisis and increasing climate-policy pressure.
• It simultaneously considers financial, organisational, ESG-related, and regional determinants of technical efficiency.
• It provides practical implications for managers, investors, and policymakers on resource allocation and energy transition strategies.

Although DEA, panel regression, ESG analyses, and regional comparisons have all been widely applied in the literature, they have typically been examined separately. Most studies focus on only one of these dimensions—technical efficiency, environmental and social responsibility performance, renewable energy companies, or regional differences—without integrating them into a single analytical framework.

Furthermore, few studies simultaneously compare renewable and traditional energy companies in Europe, the United States, and Canada during the recent period of rapid energy transition, market disruption, and increasing climate policy pressure. Existing research also rarely combines firm-level efficiency measurement with ESG performance, regional regulatory conditions, and organisational characteristics.

This study addresses these gaps by integrating DEA efficiency measurement with panel data analysis to assess both technical-operational efficiency and its determinants. It compares renewable and traditional energy companies operating in Europe and North America during the 2017–2024 period, capturing the effects of the energy transition, the energy crisis, rising energy prices, and the tightening of climate policies. Additional explanations of all abbreviations used throughout the article can be found in Appendix A.

Therefore, the novelty of this research lies not in the individual methods employed, but in their combined application to explain how technological, financial, organisational, ESG, and regional factors jointly shape firm-level technical-operational efficiency during the energy transition.

2. Review of the Literature

Technical-operational efficiency of firms refers to the ability to generate, transmit, or deliver energy while making the best possible use of resources: infrastructure, fuels, capital, technology, labour, and data. Simply put, a technically efficient company does not waste resources: it produces more energy with the same inputs or achieves the same level of production at lower costs, with fewer failures and lower emissions [1,2].

In the energy sector, technical efficiency manifests itself primarily through the high efficiency of the generating units, low grid losses, good asset availability, efficient maintenance management, grid digitalisation, and the ability to integrate renewable energy sources. In the context of the energy transition, simply “producing cheaply” is no longer enough. Flexible, low-emission and stable production is becoming increasingly important [3,4].

In Europe, technical efficiency is strongly influenced by climate policy, the ETS system, EU regulations, and increasing ESG reporting requirements. The share of renewable energy in final energy consumption increased to 25.2% in 2024, but achieving the 42.5% target by 2030 requires much faster investment pace and a profound transformation of the energy system [5]. Therefore, European energy companies are forced to modernise their networks, develop energy storage facilities, automate them, and reduce coal assets.

In the US and Canada, the transformation is more market-based and regionally diversified. In the US, growing energy demand, data centres, industrial electrification, and the development of distributed generation are key factors. Deloitte indicates that the US energy sector is entering a phase of increasing demand, digitalisation, the development of distributed energy resources (DERs), nuclear energy and the management of emissions [6]. In Canada, hydropower, grid development, the role of provinces and the need to increase the capacity of the system are key; the country assumes a strong increase in renewable energy production, including an increase in solar generation from 7 TWh in 2023 to approximately 45–65 TWh in Canada Energy Regulator (CER) scenarios [7]. The biggest difference between renewable energy companies and traditional energy companies lies in their asset structure. Renewable energy companies tend to be more flexible in terms of investment, adopt new technologies more quickly, and have a lower carbon footprint [8,9]. However, their efficiency depends on weather, location, access to the grid, storage, and quality of production forecasting. Traditional energy companies, on the other hand, have greater operational experience, stable regulated revenues, and extensive infrastructure, but are often burdened with older coal, gas, or nuclear assets. Therefore, their technical efficiency depends on their ability to modernise assets faster than the age of their existing business models [2,8]. Figure 1 presents the methodological framework of the study, including the DEA efficiency-estimation stage and the panel-regression stage used to identify the determinants of technical efficiency.

ESG plays an increasingly important role in assessing the performance of energy companies, as investors, regulators, and the public expect firms to combine sound financial performance with responsible business conduct. In the energy sector, the environmental dimension is particularly important, covering reductions in CO₂ emissions, the development of renewable energy sources, improved energy efficiency, and mitigating the negative impact of operations on the climate and natural environment [10]. Companies that invest more rapidly in low-emission technologies, modernise networks, and reduce the use of fossil fuels are perceived to be more competitive and better prepared to function in the future energy system [2,9].

Equally important is the social aspect, which in the energy sector involves, among other things, employee safety, relationships with local communities, protecting energy consumers from energy poverty, and ensuring a stable energy supply. During the energy transition, the importance of these issues has increased significantly, as the closure of conventional power plants or the development of new renewable energy investments often affect local labour markets and the social situation in industrial regions [11]. Companies that manage transformation in a socially responsible way more easily gain public acceptance and reduce the risk of conflict. The third pillar of ESG, governance, refers to the quality of corporate management, decision transparency, risk management, and regulatory compliance. This is particularly important in the energy sector due to the vast scale of investments, dependence on government policy, and high regulatory risk. Companies with a stable management model, a clear decarbonisation strategy, and transparent reporting are generally more regarded by investors and financial institutions [12].

In practice, ESG is no longer just a factor in building a company image [10]. It increasingly impacts financing costs, access to capital, and company value. Banks and investment funds are more willing to support projects aligned with climate goals, while companies delaying the transition are more vulnerable to increased operating costs, loss of investors, and regulatory pressure. In Europe, ESG is particularly important due to extensive regulations from the European Union and sustainability reporting obligations. In the US, the approaches to ESG vary regionally and politically, but there, too, the importance of low-emission investments and transparency in the operations of energy companies is growing [13].

The efficiency of energy companies is increasingly influenced by regional factors, as the energy transition is progressing differently in different parts of the world. Europe is characterised by very strong regulatory pressure stemming from the European Union’s climate policy, the ETS (Emissions Trading System) and ambitious CO₂ reduction targets. In practice, this means that European energy companies are forced to accelerate infrastructure modernisation, reduce the use of fossil fuels, and invest heavily in renewable energy sources [14]. The high prices of emission allowances make the maintenance of high-emission assets increasingly unprofitable, placing additional pressure on companies to improve their technical and organisational efficiency [15]. At the same time, companies operating in the highly regulated European energy market must reconcile economic efficiency with environmental and social compliance [16].

In the United States, the energy transition is more market-driven. Private investment, technological competition, and state-level decisions play a much more significant role [17]. As a result, energy companies operate in a more diverse regulatory environment. Some regions are actively developing renewable energy and energy storage technologies, while others continue to rely on natural gas or fossil fuels. This structure means that the efficiency of the company often depends on the ability to respond quickly to market changes, attract investors, and implement technological innovations. American energy companies are increasingly investing in grid digitisation, artificial intelligence, smart metres, and data analytics, as these technologies reduce operating costs and improve the stability of the energy system [6,18,19].

Canada, on the other hand, has a unique energy model based largely on hydropower and has a strong role for provincial governments. Access to abundant natural resources means that Canadian companies often have lower energy system emissions than companies operating in other North American countries [20]. At the same time, large geographical distances and dispersed infrastructure result in high energy transmission costs and the need to maintain highly stable transmission networks [21]. For Canadian companies, technical efficiency depends on energy production and effective infrastructure management under challenging climatic and geographic conditions [7,20,21].

In addition to regional factors, organisational factors within companies are also crucial. Modern energy is becoming an increasingly knowledge-, data- and technology-based sector, which is why companies with a culture of innovation achieve higher efficiency than those operating conservatively. Companies open to the development of new technologies implement solutions related to automation, smart grids, and energy storage more quickly, translating into lower operating costs and greater operational flexibility [22].

Efficient management of investment projects is also crucial. The transformation requires significant financial investments and the implementation of multi-year infrastructure investments [23,24]. Companies that can effectively plan investments, reduce delays, and control costs achieve higher efficiency than those with organisational problems or overly complex decision-making structures. Digital competencies among employees and managers also play an increasingly important role. The importance of data analysis, cybersecurity, grid management systems, and artificial intelligence-based technologies is growing in the energy sector, which is why companies investing in developing technological competencies are better able to cope with the growing complexity of the market [25,26,27].

Another key element of efficiency is the ability to collaborate with regulators, local governments, and local communities. In the energy sector, many investments are dependent on administrative decisions, access to public financing, and social acceptance [28]. Companies that can build good relationships with stakeholders and maintain transparent communication find it easier to implement projects and reduce the risk of social and regulatory conflicts. As a result, the efficiency of energy companies increasingly reflects a combination of technology, energy production costs, management quality, organisational adaptability, and the capacity to operate in a rapidly changing economic and political environment [11,16,29]. Technical efficiency in the energy sector is no longer only a measure of machine efficiency [2,30]. It is becoming a measure of the entire organisation’s ability to adapt. The most effective companies today are those that can simultaneously maintain supply security, reduce emissions, invest in new technologies, manage environmental and social risks, and respond to local market conditions [1,31]. In energy transformation, not necessarily the largest companies gain an advantage, but those that learn the fastest to operate in a more complex, decentralised and low-emission system [1,2,10,31,32].

The literature review demonstrates that although technical efficiency, ESG performance, renewable energy, and regional aspects of the energy transition have received considerable attention, these issues have usually been analysed independently. Existing studies rarely combine technical efficiency measurement with panel data analysis, while simultaneously accounting for ESG performance, organisational characteristics, business profile, and regional regulatory differences. Moreover, comparative evidence covering both Europe and North America during the recent energy transition period remains limited. These research gaps motivate the empirical framework adopted in this study, as shown in

Table 1. Research gap and contribution of the present study.

Previous Studies	Limitation Identified	Contribution of This Study
DEA efficiency studies	Focus mainly on technical efficiency	Combines DEA with panel regression
ESG and energy transition	Usually analyse ESG separately	Integrates ESG with DEA efficiency
Renewable energy companies	Focus only on RES firms	Compares renewable and traditional firms
Regional studies	Usually analyse one region	Compares Europe with the USA and Canada
Energy transition studies	Focus on policy or technology	Integrates financial, organisational, and regulatory determinants of efficiency
DEA applications	Mostly single-method studies	Combines DEA, panel models, ESG and regional analysis

Note: Data Envelopment Analysis (DEA).

.

3. Research Methodology

3.1. Research Sample

The empirical analysis was based on an unbalanced panel of publicly listed energy companies operating in Europe, the United States, and Canada between 2017 and 2024. Financial and ESG data were obtained from the S&P Global Market Intelligence database, ensuring consistency and comparability between firms and reporting years.

The sample included two types of energy companies: traditional electricity utilities (electric power companies) and renewable energy producers (independent power and renewable electricity producers), as classified by S&P Global Market Intelligence. Firms with incomplete financial or ESG data were excluded, resulting in a final sample of 63 companies and 438 firm-year observations. Of these, 202 observations related to European companies and 236 to firms from the United States and Canada, while 310 observations represented traditional energy companies and 128 renewable energy producers.

This sample enables comparisons across both business models and regulatory environments while maintaining a relatively homogeneous set of firms in terms of core activities. The 2017–2024 period captures major developments in the energy sector, including the acceleration of the energy transition, the energy crisis, increasing pressure on climate policies, and the growing importance of ESG reporting. The analysis begins in 2017, the first year for which sufficiently consistent ESG data were available to construct a comparable panel data set.

The sample was selected purposively, as shown in

Table 2. Characteristics of the research sample.

Characteristic	Value
Study period	2017–2024
Number of companies	63
Firm-year observations	438
Europe	202 observations
USA & Canada	236 observations
Traditional energy companies	310 observations
Renewable energy companies	128 observations
Database	S&P Global Market Intelligence

. Only companies operating throughout the analysed period and reporting complete financial and ESG information required for DEA estimation and panel regression were included. Firms with incomplete observations or missing key variables were excluded from the analysis. This procedure ensured the consistency and comparability of efficiency estimates between companies and over time.

3.2. DEA Methodology

The Data Envelopment Analysis (DEA) method, a nonparametric frontier method, was used to assess the technical efficiency of companies. DEA allows for the evaluation of the relative efficiency of decision-making units (DMUs) while simultaneously taking into account multiple inputs and outputs. The study used:

• The CCR (Charnes–Cooper–Rhodes) model, assuming constant returns to scale (CRS);
• The BCC model (Banker–Charnes–Cooper), assuming variable returns to scale (VRS);
• An output-orientated approach, assuming maximisation of outputs with given input resources.

An output-orientated DEA model was adopted because the objective of this study is to evaluate how effectively energy companies transform their existing resources into operating results. In the energy sector, reducing assets or infrastructure is rarely a feasible managerial strategy. Power plants, transmission networks, and renewable energy installations require substantial long-term investments, while financial commitments are typically tied to long-term debt and other external financing. Consequently, managers have limited flexibility to reduce resource input and instead focus on improving the performance of existing assets.

Therefore, an output-orientated specification was considered more appropriate than an input-orientated approach, as it assesses whether firms can generate higher revenues and operating profits using their existing resource base. This perspective is particularly relevant during the energy transition, when companies are expected to expand and modernise infrastructure, invest in renewable energy technologies, and strengthen grid resilience rather than reduce production capacity. DEA models were estimated using the benchmarking package in R.

The selection of inputs and outputs was guided by the resource-based view (RBV) of the firms, according to which organisational performance depends on the ability to transform strategic resources into economic results [33]. Consequently, the objective of this study is to evaluate the technical-operational efficiency rather than the engineering efficiency of electricity generation.

Total assets and total liabilities were selected as input. Total assets represent a company’s productive resource base, including generation facilities, transmission and distribution infrastructure, storage facilities, digital technologies, and other long-term assets that determine operational capacity. Total liabilities capture the financial resources supporting these assets, reflecting the long-term financing required for infrastructure development, renewable energy investments, and grid modernisation during the energy transition. Similar input specifications have been widely applied in DEA studies of firm-level efficiency in the energy sector [9,12,34].

Revenue and EBIT were selected as outputs because they capture complementary dimensions of organisational performance. Revenue reflects the firm’s ability to generate market output, whereas EBIT measures operating profitability independently of financing and taxation decisions. Together, these variables assess how effectively companies transform their asset base and financial resources into operating results.

Consequently, the DEA specification evaluates whether energy companies are able to transform their assets and financing resources into higher revenues and operating profits. In the output-orientated DEA framework, efficiency is interpreted as the ability to increase operating results while maintaining the existing level of resources.

The technical efficiency in the CCR model was defined according to the following relationship.

$${\mathit{T}\mathit{E}}_0={\displaystyle \frac{1}{\mathit{\varphi }}} \mathsf{\varphi }\ge 1$$

(1)

where $\varphi$ denotes the optimal proportional output-expansion factor for decision-making unit o. In the output-oriented CCR model, the evaluated unit itself constitutes a feasible solution at $\varphi$ = 1. Therefore, provided that the model is feasible and the input and output data satisfy the standard DEA conditions, ϕ ≥ 1, the denominator in Equation (1) cannot equal zero. A value of ${TE}_0$ = 1 indicates a technically efficient unit located on the efficiency frontier, whereas values below 1 indicate technical inefficiency.

Based on the results of the CCR and BCC models, the scale efficiency was determined, defined as

$$SE={\displaystyle \frac{CCR}{BCC}}$$

(2)

where:

• SE = 1 indicates high-scale efficiency;
• Lower values indicate the presence of scale inefficiencies.

This indicator allows us to assess the proportion of inefficiencies resulting from the suboptimal scale of a company’s operations.

To identify structural differences, comparative analyses were performed between firms.

Business Type:

• RES (RES = 1);
• Traditional energy companies (RES = 0).

Binary variables were constructed based on the sectoral classification of companies.

Region of operation:

• Europe;
• USA and Canada.

For each group, the following were calculated: mean efficiency, median, quartiles, and minimum and maximum values.

To identify the determinants of technical efficiency, variables and indicators related to the energy transition were used.

The energy transition was operationalised through:

• Classification of companies as renewable energy;
• ESG indicators;
• Environmental indicators (environmental scores);
• Regional interactions reflecting exposure to climate policy.

The direct impact of ETS2 was not modelled, as the ETS2 system was not yet operational for the period analysed. Instead, an indirect approach (regulatory exposure proxy) was used, assuming a higher exposure of European companies to climate regulations.

In the second stage of the study, panel models were estimated to analyse the determinants of technical efficiency.

The dependent variable was the DEA efficiency of the companies:

$$DEA\_Efficienc{y}_{it}$$

(3)

The following models were used: fixed effects (FE) models and dynamic panel models with lagged dependent variables.

In addition to firm fixed effects, a two-way fixed effect specification that includes year effects was estimated. Fixed effects capture common shocks that affect all companies, including the COVID-19 pandemic, the 2022 energy crisis, rising energy prices, and broader macroeconomic and regulatory changes. Their inclusion reduces the risk that economic events are incorrectly attributed to firm-specific characteristics, thereby improving the robustness of estimated relationships.

To account for the persistence of technical efficiency, a dynamic fixed-effects model with a lagged dependent variable was estimated instead of a generalised method of moments (GMM) estimator. Although GMM is commonly applied to dynamic panel models, it is generally more appropriate for panels with many cross-sectional units and a short time dimension. Given the moderate size of the present panel (63 firms observed over eight years), the GMM could generate an excessive number of instruments relative to the sample size, increasing the risk of instrument proliferation and overfitting. The dynamic fixed-effects specification was therefore considered more appropriate for the data analysed.

The general form of the model was as follows.

$$DE{A}_{it}={\beta }_0+{\beta }_{1}ES{G}_{it}+{\beta }_2{RES}_{it}+{\beta }_{3}Europ{e}_{it}+{\beta }_{4}Leverag{e}_{it}+{\beta }_{5}Profitabilit{y}_{it}+u_i+{\epsilon }_{it}$$

(4)

where:

• DEA_it—Technical efficiency of the company;
• ESG_it—ESG indicators;
• RES_it—Binary variable that defines the business profile;
• Europe_it—Regional exposure;
• Leverage_it—Debt level;
• Profitability_it—Profitability indicators;
• $u_i$—Unobservable individual effects;
• ${\epsilon }_{it}$—Random component.

The applied research approach allowed the simultaneous assessment of the technical and operational efficiency of energy companies, the identification of efficiency benchmarks and the analysis of the impact of energy transition, ESG, and regional factors on the efficiency of energy sector companies.

Panel-model diagnostics were conducted before the interpretation of the regression estimates. Multicollinearity was assessed using variance inflation factors (VIFs). Because the Total ESG Score and the Environmental Score capture closely related dimensions of corporate sustainability performance, their simultaneous inclusion was additionally examined for potential collinearity. The two indicators were subsequently entered into separate specifications: the Total ESG Score was retained in the main model, whereas the Environmental Score was included in an alternative robustness specification.

Model selection between the fixed-effects and random-effects specifications was evaluated using a regression-based robust Hausman–Mundlak test. For the preferred two-way fixed-effects specification, heteroskedasticity was examined using the Breusch–Pagan test, first-order serial correlation was assessed using the Wooldridge test for panel data, and residual normality was evaluated using the Jarque–Bera test. To account for within-firm dependence in the regression residuals, statistical inference was based on standard errors clustered at the firm level.

To assess the robustness of the DEA results, efficiency was estimated using both the CCR model assuming constant returns to scale (CRS) and the BCC model assuming variable returns to scale (VRS). The scale efficiency was then calculated as the ratio of the CCR to the BCC efficiency scores.

Comparing these measures makes it possible to assess whether the findings are sensitive to the returns-to-scale assumption. Consistent results across the CCR and BCC models indicate robustness, while scale efficiency distinguishes inefficiency arising from suboptimal operating scale from pure technical inefficiency.

This robustness assessment is particularly relevant in the energy sector, where firms differ considerably in size, ownership structure, investment intensity, and production technologies. The combined use of CCR, BCC, and scale efficiency therefore provides a more comprehensive assessment of firm-level technical efficiency than relying on a single DEA specification. The adopted approach is consistent with previous DEA studies in the energy sector, where financial variables are commonly used as proxies for organisational resources and operational performance [35,36,37].

4. Results

The results indicate that the technical efficiency of the energy companies varied significantly between regions and between renewable energy companies and traditional energy companies. The comparison between the CCR and BCC models also serves as a robustness assessment of the DEA results. Consistently higher efficiency scores obtained under the BCC specification indicate that a substantial proportion of inefficiency is attributable to scale effects rather than to pure technical inefficiency. Consequently, the main conclusions remain stable across alternative DEA specifications, supporting the robustness of the empirical findings. As presented in

Table 3. Comparison of CCR, BCC, and scale efficiency scores (robustness assessment of DEA results).

Year	N	CCR Mean	CCR Median	BCC Mean	BCC Median	Scale Mean	Efficient CCR	Efficient BCC
2017	55	0.5719	0.5240	0.6755	0.7143	0.8469	6	10
2018	54	0.5218	0.4381	0.6289	0.5902	0.8239	7	11
2019	54	0.5749	0.5133	0.6871	0.6604	0.8346	9	11
2020	55	0.5364	0.4786	0.6191	0.5853	0.8659	7	12
2021	58	0.4058	0.3197	0.6040	0.5872	0.6703	6	13
2022	56	0.4086	0.3370	0.5653	0.5435	0.7528	5	10
2023	54	0.4233	0.3284	0.6113	0.5872	0.6907	6	12
2024	52	0.4235	0.3428	0.5721	0.5435	0.7618	5	9

Notes: N—number of firm-year observations; CCR—Charnes–Cooper–Rhodes efficiency; BCC—Banker–Charnes–Cooper efficiency; Scale—scale efficiency calculated as CCR/BCC. Source: Author’s calculations based on research data.

, the average technical efficiency measured by the CCR model in the analysis period 2017–2024 was approximately 0.48, while the average efficiency of the BCC model reached approximately 0.62. This means that after accounting for variables-scale effects, companies operated significantly more efficiently, suggesting a significant impact of scale on energy sector performance. At the same time, the average scale efficiency was approximately 0.78, indicating that a significant portion of the companies’ inefficiency was due to inadequate operating scale. A significant decline in the average efficiency of CCR was observed between 2021 and 2023, which may be attributed to the growing volatility of the energy market, the increasing financing costs, and the effects of the energy transition and the energy crisis in Europe. It is worth emphasising that the number of companies on the efficiency frontier was relatively small: only a few companies achieved an efficiency of 1, confirming the high competitiveness of the sector and significant differences in companies’ ability to efficiently use resources.

To further assess the robustness of the DEA results, a sensitivity analysis was performed using an alternative input specification. Although the baseline model included total assets and total liabilities as input, the alternative specification excluded liabilities, retaining total assets as the only input. In both models, total revenue and EBIT were used as output.

Excluding liabilities reduced the average efficiency scores, with the CCR efficiency decreasing from 0.4828 to 0.3209, the BCC efficiency from 0.6204 to 0.5448, and the scale efficiency from 0.7802 to 0.6275. This indicates that the alternative specification produces a more restrictive efficiency frontier.

Despite lower absolute efficiency scores, the main findings remained unchanged. Traditional energy companies continued to outperform renewable energy companies (average CCR efficiency of 0.3259 versus 0.3089), while European firms remained more efficient than companies from the United States and Canada (0.3370 versus 0.3072).

Overall, the sensitivity analysis confirms that the study’s conclusions are robust to the choice of DEA input specification. Although excluding liabilities affects the absolute level of efficiency scores, it does not alter the relative differences between business models or regions, supporting the interpretation that these differences primarily reflect firm characteristics and operating conditions rather than the specification of the model, as shown in

Table 4. Sensitivity analysis of DEA results using an alternative input specification.

Specification	N	CCR Mean	BCC Mean	Scale Mean
Baseline DEA: Assets + Liabilities → Revenue + EBIT	438	0.4828	0.6204	0.7802
Sensitivity DEA: Assets → Revenue + EBIT	438	0.3209	0.5448	0.6275
Difference	438	−0.1619	−0.0755	−0.1526

Notes: N—number of firm-year observations; CCR—Charnes–Cooper–Rhodes efficiency; BCC—Banker–Charnes–Cooper efficiency; Scale—scale efficiency calculated as CCR/BCC. Source: Author’s calculations based on research data.

.

The analysis presented in

Table 5. Comparison of the efficiency of renewable energy companies and traditional energy companies.

RES	Specialization	N	CCR Mean	CCR Median	BCC Mean	Scale Mean	ESG Mean	Env Mean
0	Electric Power Companies	310	0.5178	0.4440	0.6440	0.8088	48.9968	51.6903
1	Independent Power and Renewable Electricity Producers	128	0.3980	0.3337	0.5630	0.7108	51.4844	54.8594

Notes: RES—renewable energy producer indicator; ESG—environmental, social, and governance score; Env—environmental score; N—number of firm-year observations. Source: Author’s calculations based on research data.

indicates that renewable energy companies achieved, on average, lower technical efficiency than traditional energy companies. The average CCR efficiency for traditional energy companies was higher than for renewable energy companies, as was their scale efficiency. These results suggest that despite the dynamic development of the renewable energy sector, renewable energy companies continue to operate on a suboptimal scale and incur higher costs of technological transformation. At the same time, renewable energy companies had a higher average ESG and environmental scores, indicating their greater commitment to sustainable development and climate policy. This indicates that renewable energy companies are more environmentally advanced, but an ESG advantage does not automatically translate into higher economic efficiency.

The results presented in

Table 6. Comparison of the efficiency of European companies with those of the US and Canada.

Region	N	CCR Mean	CCR Median	BCC Mean	Scale Mean	ESG Mean	Env Mean
Europe	202	0.5153	0.4183	0.6602	0.7767	56.1139	58.7673
United States and Canada	236	0.4551	0.3970	0.5863	0.7831	44.2542	47.3517

Notes: RES—renewable energy producer indicator; ESG—environmental, social, and governance score; Env—environmental score; N—number of firm-year observations. Source: Author’s calculations based on research data.

show that European companies achieved, on average, higher technical efficiency than companies from the United States and Canada. The average CCR efficiency for Europe was higher than for North American companies, while at the same time, the scale efficiency was very similar. Importantly, European companies achieved significantly higher values of the ESG and Environmental Score, which may indicate greater regulatory maturity and stronger climate pressures in Europe. These results suggest that European energy companies are more advanced in the energy transition process and are more aligned with climate policy requirements than North American companies.

Particularly interesting results are presented in

Table 7. Efficiency of the company by region and type of business.

Region	RES	N	CCR Mean	CCR Median	BCC Mean	Scale Mean	ESG Mean	Env Mean
Europe	0	126	0.5682	0.4782	0.6950	0.8100	56.1746	58.3810
Europe	1	76	0.4274	0.3614	0.6024	0.7214	56.0132	59.4079
United States and Canada	0	184	0.4833	0.4305	0.6092	0.8079	44.0815	47.1087
United States and Canada	1	52	0.3550	0.3225	0.5053	0.6954	44.8654	48.2115

Notes: RES—renewable energy producer indicator; ESG—environmental, social, and governance score; Env—environmental score; N—number of firm-year observations. Source: Author’s calculations based on research data.

, which combines the criteria for the region and type of business. The highest efficiency was achieved by traditional European energy companies, while the lowest results were achieved by renewable energy companies from the United States and Canada. This indicates that the development of renewable energy sources alone does not guarantee an efficiency advantage; regulatory experience, access to infrastructure, and the scale of a company’s operations also play a role. Despite their high exposure to climate policy, traditional European companies demonstrated relatively high operational efficiency, which may indicate more effective adaptation to the conditions of the energy transition.

The results of the panel model presented in

Table 8. Results of the panel models.

Model	Term	Coef	std_err	p_Value
Pooled OLS	const	0.4870	0.0360	0.0000
	ESG_z	0.1046	0.0903	0.2466
	env_z	−0.1396	0.0881	0.1130
	RES	−0.0747	0.0745	0.3163
	europe	0.0523	0.0696	0.4526
	RES_x_europe	−0.0371	0.1003	0.7116
	ln_assets_z	0.0128	0.0246	0.6038
	leverage_z	−0.0567	0.0252	0.0244
	ROA_z	0.1402	0.0261	0.0000
Firm FE	const	0.4661	0.0061	0.0000
	ESG_z	0.0633	0.0764	0.4069
	env_z	−0.0646	0.0731	0.3773
	RES	0.0682	0.0219	0.0019
	europe	0.0290	0.0165	0.0787
	RES_x_europe	−0.5747	0.0433	0.0000
	ln_assets_z	−0.3714	0.0436	0.0000
	leverage_z	−0.0043	0.0148	0.7737
	ROA_z	0.0929	0.0272	0.0006
Two-way FE	const	0.5357	0.0213	0.0000
	ESG_z	0.0197	0.0679	0.7714
	env_z	−0.0078	0.0646	0.9041
	RES	0.0109	0.0312	0.7262
	europe	0.0265	0.0128	0.0386
	RES_x_europe	−0.3287	0.0925	0.0004
	ln_assets_z	−0.1707	0.0881	0.0527
	leverage_z	−0.0445	0.0171	0.0094
	ROA_z	0.1172	0.0284	0.0000
Dynamic fixed-effects model	const	0.5122	0.0268	0.0000
	lag_eff_z	−0.0451	0.0269	0.0937
	ESG_z	0.0487	0.0830	0.5575
	env_z	−0.0328	0.0749	0.6610
	RES	−0.0049	0.0341	0.8859
	europe	0.0265	0.0161	0.0996
	RES_x_europe	−0.3969	0.0931	0.0000
	ln_assets_z	−0.2075	0.0911	0.0228
	leverage_z	−0.0662	0.0187	0.0004
	ROA_z	0.1062	0.0289	0.0002

Notes: OLS—ordinary least squares; FE—fixed effects; RES—renewable energy producer indicator; ESG—environmental, social, and governance score; ROA—return on assets; coef—estimated coefficient; std_err—clustered standard error; the suffix _z denotes a standardised variable. Source: Author’s calculations based on research data.

show that the relationships between technical efficiency and ESG indicators, business profile and region were not statistically clear. Panel estimations indicate that leverage had a statistically significant negative effect on technical efficiency (β = −0.0445, p = 0.009 in the two-way FE model), while profitability exerted a positive effect (β = 0.1172, p < 0.001). Size was negatively associated with efficiency (β = 0.1707, p = 0.053), suggesting that larger companies do not necessarily operate more efficiently after controlling for firm-specific effects. This may suggest that the effects of environmental investments are long-term and do not immediately translate into improved economic efficiency for energy companies.

The diagnostic results for the panel regression models are presented in

Table 9. Multicollinearity and specification diagnostics linked to the panel models in Table 8.

Diagnostic	Statistic	p-Value	Conclusion
Initial: VIF—Total ESG Score	25.363	—	Problematic multicollinearity
Initial: VIF—Environmental Score	22.617	—	Problematic multicollinearity
Model A: VIF—Total ESG Score	1.633	—	No problematic multicollinearity
Model A: VIF—RES	2.423	—	No problematic multicollinearity
Model A: VIF—Europe	1.606	—	No problematic multicollinearity
Model A: VIF—RES × Europe	2.989	—	No problematic multicollinearity
Model A: VIF—ln(assets)	1.621	—	No problematic multicollinearity
Model A: VIF—leverage	1.020	—	No problematic multicollinearity
Model A: VIF—ROA	1.104	—	No problematic multicollinearity
Robust Hausman–Mundlak test	χ2(4) = 78.171	<0.001	Fixed effects preferred
Breusch–Pagan test	χ2(4) = 6.907	0.141	Homoskedasticity not rejected
Wooldridge test	F(1,374) = 5.431	0.020	First-order serial correlation detected
Jarque–Bera test	169.328	<0.001	Residual normality rejected

Notes: The Total ESG Score and Environmental Score were included jointly in the models reported in Table 8. Their high VIF values indicate substantial multicollinearity; therefore, the corresponding coefficients should be interpreted with caution. VIF—variance inflation factor; RES—renewable energy producer indicator; ROA—return on assets. Standard errors were clustered at the firm level. Source: Author’s calculations based on research data.

. The simultaneous inclusion of the Total ESG Score and the Environmental Score resulted in relatively high VIF values (25.363 and 22.617, respectively), indicating substantial multicollinearity between these two sustainability indicators. For all remaining explanatory variables, VIF values ranged from 1.020 to 2.989, suggesting that multicollinearity was not a material concern. The remaining diagnostic tests reported in Table 9 refer to the panel models presented in Table 8.

The robust Hausman–Mundlak test rejected the null hypothesis underlying the random-effects specification, χ²(4) = 78.171, p < 0.001, supporting the use of the fixed-effects approach. For the preferred two-way fixed-effects model, the Breusch–Pagan test did not provide evidence of heteroskedasticity, χ²(4) = 6.907, p = 0.141. However, the Wooldridge test indicated statistically significant first-order serial correlation, F(1,374) = 5.431, p = 0.020. The Jarque–Bera test also rejected the assumption of residual normality, JB = 169.328, p < 0.001. The residuals were moderately negatively skewed and exhibited excess kurtosis. In view of the detected serial correlation and the panel structure of the data, all reported statistical inference is based on standard errors clustered at the firm level.

Statistical tests, presented in

Table 10. Results of statistical tests for the efficiency of energy companies.

Test	Group 0	Group 1	Mean 0	Mean 1	t_stat	p_Value	n0	n1
eff_CCR_output by RES	0	1	0.5178	0.3980	4.5438	0.0000	310	128
eff_BCC_output by RES	0	1	0.6440	0.5630	2.8936	0.0042	310	128
scale_efficiency by RES	0	1	0.8088	0.7108	4.2073	0.0000	310	128
eff_CCR_output by Europe	0	1	0.4551	0.5153	−2.3209	0.0208	236	202
eff_BCC_output by Europe	0	1	0.5863	0.6602	−2.8807	0.0042	236	202
scale_efficiency by Europe	0	1	0.7831	0.7767	0.3062	0.7596	236	202

Notes: t_stat—t-statistic; p_value—p-value; n0 and n1—numbers of observations in the compared groups. Source: Author’s calculations based on research data.

, were also an important element of the study. The results of the mean difference tests confirmed statistically significant differences between renewable energy companies and traditional energy companies for both CCR and BCC efficiency, as well as scale efficiency. Similarly, significant differences were found between European and North American companies. This indicates that both the company’s business profile and the region of operation have a significant impact on the level of technical efficiency of energy sector companies.

Furthermore, the ranking analysis presented in

Table 11. The most and least efficient companies in the energy sector in 2023.

Company	Region	Specialization	Assets	Revenue	EBIT	Total Debt	ESG	Env	Climate	Emissions Targets	Gov	Social	RES
TER0 ENERGY Industrial Commercial Technical Societe Anonyme	Europe	Electric Power Companies	2,012,872	323,004	142,895.1	1,141,120	48	57	46	64	41	44	0
Admie Holding S.A.	Europe	Electric Power Companies	849,504.2	41,865.66	41,154.81	39.80439	18	16	48	16	12	0	0
VERBUND AG	Europe	Independent Power and Renewable Electricity Producers	19,653,548	5,662,478	1,348,832	3,863,082	59	69	66	100	51	54	1
TAURON Polska Energia S.A.	Europe	Electric Power Companies	9,937,018	6,519,959	587,068.4	3,245,803	33	34	31	20	34	32	0
Fortis Inc.	United States and Canada	Electric Power Companies	43,541,830	6,668,614	1,894,976	19,519,699	40	31	45	54	52	39	0
Enerjisa Enerji A.S.	Europe	Electric Power Companies	2,373,221	3,815,947	520,112.5	855,276.1	51	67	100	90	30	0	0
The AES Corporation	Europe	Electric Power Companies	32,963,000	11,141,000	2,566,000	19,013,000	74	76	82	74	84	62	0
Hawaiian Electric Industries. Inc.	Europe	Electric Power Companies	15,822,637	2,850,379	391,914	2,601,000	42	40	61	35	47	40	0
Northland Power Inc.	United States and Canada	Electric Power Companies	10,171,409	1,670,054	620,830.3	6,429,262	55	61	62	40	46	54	0
Enel SpA	Europe	Electric Power Companies	2.35 × 108	98,434,447	11,088,557	82,501,990	88	90	75	65	84	90	0
TransAlta Corporation	United States and Canada	Electric Power Companies	7,290,457	2,170,886	324,715.4	3,162,411	47	55	74	50	45	38	0
Otter Tail Corporation |	United States and Canada	Electric Power Companies	2,754,830	1,196,844	247,692	874,637	30	27	37	64	41	24	0
Edison Inter0tio0l	United States and Canada	Electric Power Companies	74,745,000	14,905,000	3,158,000	29,533,000	44	45	47	60	45	40	0
Duke Energy Corporation	Europe	Independent Power and Renewable Electricity Producers	1.7 × 108	24,201,000	5,969,000	68,263,000	59	61	61	35	62	53	1
Ormat Technologies. Inc.	United States and Canada	Electric Power Companies	4,425,678	663,084	177,724	1,934,573	38	45	27	0	38	28	0
Brookfield Renewable Partners L.P.	United States and Canada	Independent Power and Renewable Electricity Producers	55,867,000	4,104,000	950,000	21,993,000	51	58	84	60	55	38	1
RWE Aktiengesellschaft	Europe	Independent Power and Renewable Electricity Producers	1.62 × 108	29,062,072	1,935,027	20,236,552	70	70	81	100	79	61	1
Avangrid. Inc.	Europe	Independent Power and Renewable Electricity Producers	37,823,000	6,320,000	816,000	11,359,000	76	75	73	94	78	75	1
PG&E Corporation	United States and Canada	Independent Power and Renewable Electricity Producers	1.03 × 108	20,642,000	3,164,000	46,168,000	31	28	60	80	36	30	1
ERG S.p.A.	United States and Canada	Electric Power Companies	6,827,981	711,376.7	209,731.6	4,032,882	71	85	81	50	54	70	0
TransAlta Renewables Inc.	United States and Canada	Independent Power and Renewable Electricity Producers	2,962,489	374,978.5	68,613.08	775,193.8	34	39	100	34	22	41	1
Boralex Inc.	Europe	Electric Spólek energetycznych	4,544,485	551,298.1	148,395.7	3,115,789	54	57	40	0	59	45	0
ALLETE. Inc.	Europe	Independent Power and Renewable Electricity Producers	6,422,300	1,419,200	157,400	1,993,800	29	28	34	39	38	24	1
Exelon Corporation	Europe	Electric Power Companies	1.33 × 108	17,938,000	2,872,000	34,855,000	71	72	83	100	70	69	0
Portland General Electric Company	United States and Canada	Electric Power Companies	9,494,000	2,396,000	373,000	3,600,000	36	37	39	50	38	33	0
Fortum Oyj	United States and Canada	Electric Power Companies	1.7 × 108	7,595,809	1,690,192	1,9747,526	55	60	78	50	49	54	0
BKW AG	United States and Canada	Electric Power Companies	13,251,390	3,780,431	370,866.4	2,259,848	31	29	5	0	40	27	0
IDACORP. Inc.	Europe	Electric Power Companies	7,210,515	1,458,084	314,402	2,000,640	33	35	41	43	38	25	0
Public Power Corporation S.A	United States and Canada	Electric Power Companies	20,219,702	6,749,402	247,002.2	5,919,294	40	43	47	48	31	45	0
Encavis AG	United States and Canada	Electric Power Companies	3,657,327	393,514.2	110,476.3	1,850,321	39	45	14	40	38	32	0

Notes: Assets—total assets; Revenue—total revenue; EBIT—earnings before interest and taxes; Total debt—total debt reported by the company; ESG—total environmental, social, and governance score; Env—environmental score; Climate—climate-related performance score; Emissions targets—score reflecting the company’s emissions-reduction targets; Gov—governance score; Social—social score; RES—renewable energy producer indicator, coded as 1 for independent power and renewable electricity producers and 0 for traditional electric power companies. Higher ESG-related scores indicate stronger performance in the respective assessment category. Source: Author’s calculations based on research data.

identified companies that serve as efficiency benchmarks for the sector. Among the most efficient companies were both traditional energy companies and entities that implement the energy transition, indicating that high efficiency is achievable regardless of the business model provided proper resource management and an optimal scale of operations.

5. Discussion

The discussion is organised according to the five research hypotheses. Each subsection relates the empirical findings to the corresponding DEA or panel-model results and to previous studies. Regional findings are additionally interpreted in the context of the different policy and institutional environments of Europe, the United States, and Canada.

5.1. Business Profile and Technical Efficiency—Hypothesis H1

As shown in Table 5, Table 7 and Table 10, technical efficiency varied according to business profile and operating scale. Traditional energy companies achieved higher average CCR and BCC efficiency than renewable energy companies, and the differences were statistically significant. These results are consistent with previous studies highlighting the operational challenges associated with periods of intensive investment and transformation [29,30,32]. Consequently, Hypothesis H1, which predicted higher efficiency among renewable energy companies, was not supported.

Although this outcome may seem counterintuitive given the growing importance of renewable energy, it can be largely explained by structural differences between the two groups. Traditional utilities typically benefit from longer market experience, more developed infrastructure, larger operating scale, and more stable financial flows, allowing them to exploit economies of scale more effectively. On the contrary, renewable energy companies are still undergoing rapid expansion and are investing heavily in new technologies and infrastructure, which can temporarily reduce technical efficiency [1,12,14,28].

As expected, renewable energy companies achieved higher environmental and ESG scores, confirming their stronger commitment to sustainability and climate objectives [9,16,23]. However, the superior environmental performance did not translate into higher technical efficiency. The capital-intensive nature of the energy transition and the long repayment period of renewable energy investments mean that lower DEA scores are more likely to reflect the current stage of development than weaker long-term competitiveness.

5.2. Regional Efficiency and Local Policy Context—Hypothesis H2

Regional differences were also evident in Table 6, Table 7 and Table 10. European companies recorded significantly higher average CCR efficiency than firms from the United States and Canada (0.515 vs. 0.455; t = 2.32, p = 0.021), supporting Hypothesis H2. One possible explanation for this is the more advanced climate policy and regulatory framework implemented within the European Union [5,23,24]. Long-term exposure to EU ETS, decarbonisation policies, and stricter environmental regulations has probably encouraged earlier adaptation of business models, more efficient asset use, and greater operational discipline [17,23,24]. Although these regulations increase compliance costs, they can also accelerate organisational modernisation and efficiency improvements.

The institutional context in the United States differs substantially from the European framework. The US energy transition is more decentralised and market-driven, with energy regulation, renewable-energy targets, electricity-market design, and investment incentives varying across individual states. Consequently, energy companies may operate under considerably different policy conditions even within the same national market. Private investment, technological competition, access to financing, digitalisation, and the ability to respond rapidly to changes in energy demand therefore play a particularly important role in shaping operational efficiency [6,17,18,19]. This regulatory heterogeneity may partly explain why the North American results were less uniform than those observed for European companies.

Canada represents a different policy and infrastructural environment. Provincial governments play a central role in electricity regulation, and the energy mix differs considerably across provinces. Hydropower provides a substantial share of electricity generation in several regions, while long transmission distances, dispersed settlements, climatic conditions, and the need to maintain reliable supply in remote areas increase infrastructure and network-management costs [7,20,21]. These characteristics may affect the technical efficiency of Canadian companies differently from that of US companies.

However, the United States and Canada were combined into a single North American category in the empirical analysis. The country-specific policy discussion therefore provides an institutional interpretation of the results rather than separate econometric evidence for the two countries.

5.3. ESG Performance and Technical Efficiency—Hypothesis H3

The panel-regression results presented in Table 8 paint a more nuanced picture of the relationship between ESG performance and technical efficiency. In all specifications, including the preferred two-way fixed effects model, neither the overall ESG score (β = 0.0197, p = 0.771) nor the Environmental Score (β = −0.0078, p = 0.904) was significantly associated with technical efficiency. This finding is consistent with the mixed evidence reported in previous studies [9,15,16]. However, because neither ESG indicator was statistically significant, the empirical results do not support Hypothesis H3.

5.4. Scale Efficiency—Hypothesis H4

The results presented in Table 3, Table 5 and Table 10 support Hypothesis H4 concerning the role of operating scale. Average scale efficiency amounted to 0.780, indicating that approximately 22% of the potential efficiency gap was associated with operation at a non-optimal scale. Renewable energy companies recorded lower scale efficiency than traditional firms (0.711 versus 0.809), which is consistent with earlier studies identifying scale-related inefficiencies in the renewable energy sector [14,29,30]. These findings suggest that many renewable energy companies have not yet reached the operating scale required to fully exploit economies of scale. The inclusion of year effects accounted for common external shocks, including the COVID-19 pandemic, the 2022 energy crisis, fluctuations in energy prices, and changes in the regulatory environment, without materially altering the main conclusions. Hypothesis H4 was therefore supported.

The absence of a significant ESG effect should not be interpreted as evidence that sustainability initiatives lack economic value. Rather, it reflects the different timings of investment costs and economic benefits. Energy companies must first finance renewable energy projects, grid modernisation, digital technologies, and emission reduction, while improvements in productivity, asset use, operating costs, and regulatory resilience typically emerge only over a longer period of time [10,11,28]. Similar conclusions have been reported in studies showing that ESG investments are more likely to improve performance in the long term than in the short term, particularly in capital-intensive industries such as energy [9,15,16].

5.5. Financial and Organisational Determinants—Hypothesis H5

Firm-level financial characteristics, unlike ESG indicators, showed a clear relationship with technical efficiency. As reported in Table 8, profitability was positively associated with efficiency (β = 0.1172, p < 0.001), whereas leverage was negatively associated with efficiency (β = −0.0445, p = 0.009). These findings provide partial support for Hypothesis H5 and suggest that the energy transition influences firm performance primarily through financial, organisational, and scale-related factors [1,2,12,28]. Although ESG performance reflects companies’ commitment to sustainability, the environmental component of the hypothesis was not statistically confirmed. Therefore, successful energy transition strategies should combine renewable energy investment with effective resource management, financial resilience, an appropriate operating scale, and organisational adaptation to changing market and regulatory conditions.

5.6. Robustness, Synthesis, and Practical Implications

In general, the empirical findings indicate that Hypotheses H2 and H4 were supported, whereas Hypotheses H1 and H3 were not supported. Hypothesis H5 received partial support, as the financial and organisational determinants were statistically relevant, while the environmental component of the hypothesis was not confirmed.

The robustness assessment showed that the principal findings remained stable across the static and dynamic specifications and the alternative DEA model. The panel diagnostics nevertheless indicated first-order serial correlation and non-normal residuals. Statistical inference was therefore based on standard errors clustered at the firm level, as reported in the Results Section. Overall, H2 and H4 were supported, H1 and H3 were not supported, and H5 received partial support.

6. Conclusions

The findings suggest that the technical efficiency of energy companies is shaped by a combination of technological profile, the share of renewable energy sources (RES) in the business model, organisational characteristics, operating scale, financial structure, and adaptability to changing market and regulatory conditions. Empirical analysis indicates that traditional energy companies achieved a higher average technical efficiency than renewable energy companies during the analysis period [29,30]. However, this should not be interpreted as evidence that the renewable energy sector is inherently less efficient or less competitive. Rather, it may reflect the fact that many renewable energy companies remain in a phase of intensive expansion and continue to incur substantial investment costs associated with the energy transition [1,14]. On the contrary, traditional utilities generally benefit from a longer operating experience, a more developed infrastructure, and greater financial stability.

Renewable energy companies also recorded higher environmental and ESG scores, indicating a stronger commitment to sustainability and climate-related objectives [9,16]. However, these higher environmental scores were not directly associated with greater technical efficiency. This finding suggests that improving technical efficiency in the renewable energy sector requires a combination of technological investment, organisational development, more effective resource allocation, and, where appropriate, a larger operating scale.

The results also indicate that European companies achieved, on average, higher technical efficiency than companies operating in the United States and Canada [5,23,24]. The regional effect remained statistically significant in the two-way fixed effect model (β = 0.0265, p = 0.039), suggesting that firms operating in Europe tended to achieve slightly higher technical efficiency after controlling for firm characteristics. This pattern may be associated with long-term regulatory pressure related to climate policy and decarbonisation, which may have encouraged earlier organisational adaptation and more efficient resource use. These results should be interpreted as evidence of association rather than causality. In practical terms, the findings suggest that well-designed climate policies can support organisational modernisation and more efficient resource management without necessarily reducing the competitiveness of the energy sector.

Another important finding concerns the operating scale. The results suggest that many companies, particularly renewable energy companies, have not yet reached the scale required to fully benefit from economies of scale [14,30]. This may indicate that improving technical efficiency in the renewable energy sector depends not only on additional technological investment but also on organisational development, more effective resource allocation, and, where appropriate, greater operational scale.

The results further suggest that the relationship between ESG performance and technical efficiency is more complex than often assumed [9,15,16]. Although the ESG indicators were not statistically significant, they were also not associated with lower technical efficiency. Instead, profitability exhibited the strongest positive association with efficiency (β = 0.1172, p < 0.001), whereas leverage was negatively associated with efficiency (β = 0.0445, p = 0.009). These findings may indicate that the benefits of sustainability investments are realised over a longer period and may become more apparent as the energy transition progresses [10,11,28].

The study also offers several practical implications. For managers, the results may help identify areas where operational efficiency could be improved. For investors, they may provide additional information when assessing firms’ resilience to regulatory and market changes associated with the energy transition. For policymakers, the findings suggest that supporting the energy transition requires not only promoting renewable energy deployment but also creating conditions that facilitate efficient resource management, financial stability, and an appropriate operating scale.

The robustness of the empirical findings was supported by the DEA sensitivity analysis, in which liabilities were excluded from the input specification. Consistent results obtained under alternative DEA specifications, together with the two-way fixed-effects model controlling for common year-specific shocks, suggest that the principal conclusions are not driven by a particular modelling approach. Although the alternative specification produced lower absolute efficiency scores, the relative differences between business models and regions remained unchanged, providing additional confidence in the robustness of the reported findings.

In general, the findings suggest that the energy transition alone is unlikely to guarantee higher technical efficiency. Instead, technical efficiency appears to be associated with the ability of companies to combine technological investment with effective financial management, organisational capabilities, and an appropriate operating scale. Therefore, these factors may play an important role in shaping the future competitiveness of energy companies operating under increasingly demanding regulatory and market conditions.