Corporate Governance Implications for Sustainable Performance: Focus on Leading Energy Producers in Denmark, Estonia, Latvia, Lithuania, and Sweden

Abstract

This paper aims to evaluate corporate governance in relation to enterprise performance indicators in order to enhance it. The intention is not only to align with the interests of shareholders, but also to foster competitive, sustainable, and inclusive growth. For this purpose, the leading energy producer in each of the five countries—Denmark, Estonia, Latvia, Lithuania, and Sweden—was investigated to evaluate their corporate governance performance. An analysis was conducted, employing regression analysis, Pearson correlation, and descriptive statistics. The influence of corporate governance on the performance of chosen enterprises was examined, utilising specifically developed models. The findings reveal that the corporate governance variables are diverse, and financial metrics exhibit significant variability, reflecting the complexity of the energy industry. The research results confirm that larger and more varied boards positively impact the performance of state-owned power suppliers and increase their net income. The presence of independent members was also found to contribute to the net income growth of state-owned power suppliers. However, the study indicated that the frequency of audit meetings does not necessarily increase earnings. Still, larger audit committees can contribute to CG decision-making processes concerning debt management. The results also implied the need to consider the qualifications of the board members and its composition for proper power interruption management to minimise the frequency and duration of power outages. Therefore, it must be of pivotal focus for respective corporate governance duties. In this respect, the need for more specific and regular assessments was also found to be justified regarding industry-specific challenges related to power system disruptions. Customer-centric strategies should deserve relevant attention as well. The enforcement of the management audit function could be a solution. Consequently, assessing the governance structures and decision-making processes must be systematic for energy producers due to the business dynamics leading to the revaluation of the evolving challenges and possible solutions aimed at the competitive and sustainable development of the energy sector.

1. Introduction

In the dynamic and complex energy industry landscape, businesses face numerous challenges, such as rapidly evolving consumer demands, emerging technologies, and constantly changing regulations. To navigate these challenges and remain competitive, it is essential to establish corporate governance that can effectively manage the intended objectives and related risks and ensure compliance with regulatory requirements. In this respect, enabling businesses to adapt quickly to changing business conditions and to program the latter requires corporate governance to be aligned with enterprise performance. Consequently, the purpose, structure, and performance of corporate governance have been discussed extensively by scholars [1,2] over the last decade, seeking to characterise the implications of corporate governance for the competitiveness of business undertakings. Various solutions [3,4] related to decision-making and performance measurement based on different sets of key performance indicators have been explored [5]. However, they are patchy in addressing the specifics of economic sectors.

Furthermore, scholars have addressed the challenges of social responsibility and sustainability reporting regarding corporate governance and business performance [6]. Scientists have also applied a holistic approach to long-term economic, social, and environmental sustainability in relation to the use of enterprises’ assets [7]. However, these studies are also scarce regarding particulars of the industries and concerning privately owned and state-owned enterprises. Within this context, this paper seeks to evaluate the influence of corporate governance on business operations, aiming to enhance decision-making processes in terms of strategic, tactical, and operational management objectives for state-owned power suppliers. This decision was based on the limited scholarly research available on this subject. For instance, as of 4 July 2024, only 102 papers could be found on the Web of Science and 51 in the Scopus database associated with state-owned organisations in the energy sector, of which only a few could be (partially) directly linked to the theme.

Furthermore, the power suppliers are tasked with delivering reliable and sustainable energy services and face complex challenges when aiming for the efficiency and effectiveness of their operations. Entities must commit to profitability objectives, be fiscally responsible, and ensure customer satisfaction when serving their primary needs. In this regard, challenges emerge when assessing corporate governance performance in terms of measuring business performance in energy generation and supply. Existing governance assessment methodologies [5] often lack the specificity required to address the intricate dynamics of state-owned energy enterprises, especially for the sustainable provision of energy generation and supply-related services. The latter hampers strategic decision-making and obstructs energy producers from leveraging their potential to navigate the evolving opportunities in the energy market and ensure reliable services. This paper aims to address these gaps. Therefore, the author conducted an experiment to investigate the corporate governance performance of the primary state-owned energy producer in each of the five countries: Denmark, Estonia, Latvia, Lithuania, and Sweden. The selected enterprises were chosen due to the governments’ active use of public investments to transition their energy sectors towards a greater reliance on renewable energy-based solutions in alignment with the EU Green Deal [8].

Additionally, the integration of the energy markets of the Baltic countries with the Scandinavian electricity market to ensure sustainable power supply [9] has been facilitated by the implementation of five regulatory energy packages aimed at liberalising the energy market in the European Union (EU) [10]. The resulting reconfiguration of power supply networks inherently necessitates the re-evaluating of business strategy and related business models. Consequently, the corporate governance performance of the enterprises must be adaptable and proactive in response to regulatory changes. The research objective is to assess, using a set of regression models, the effect of corporate governance variables on selected key performance indicators (KPIs) of selected power suppliers, testing respectively formulated hypotheses. The models compare the selected KPIs against the relevant corporate governance variables, considering energy generation and supply specifics. As a result, the proposed corporate governance assessment framework can be regarded as a tool for assessing and improving governance structures, ensuring a coherent integration of operational efficacy, financial responsibility, and corporate governance of energy producers.

This article is organised into six sections. Section 2 focuses on the assessment approaches of corporate governance, its implications for entities’ performance, and the specifics of performance measurement in the energy sector. Section 3 reveals the proposed methodology for assessing the impact of corporate governance on enterprise performance variables of five selected power suppliers. Section 4 and Section 5 discuss the outcomes of the solution applied regarding the implications of corporate governance on the performance of the investigated energy producers and the resulting rationale for strategic management. The final section presents conclusions and insights for future research.

2. Literature Review

2.1. Understanding and Measuring the Performance of Corporate Governance

Corporate governance (CG) pertains to policies, procedures, and mechanisms that supervise an organisation’s management, direction, and control. It is worth noting that as a term, CG came into focus in the 1970s in the United States [11], while in Europe, it underwent scrutiny in the 1990s [12].

CG is intended to uphold fairness, transparency, responsibility, and accountability within the organisation as a framework and ensure adherence to all the requirements. However, CG must exhibit flexibility and proactivity in the dynamic business environment. This involves adapting to changes and having the foresight and capability to promptly implement responsive measures to ensure the ongoing competitiveness and sustainable development of the pertinent undertakings. It is essential to consider the sustainability of businesses, not only in meeting their current needs without compromising the ability of future generations to meet their own needs [13] but also in compliance with corporate social responsibility norms [14]. This means that businesses should enhance society in every interaction while striving to achieve their business objectives.

In such a context, organisational inertia must be a pivotal focus for CG’s efficiency and effectiveness, as besides sustainability objectives, business organisations are required to be profitable and competitive in the short and long run. This applies to private and state-owned organisations. Inertia adversely affects the business performance indicators. It denotes resource management inflexibility, and inflexibility in adjusting organisational processes and procedures, lowering the speed of adaptability to a changing business environment [15,16]. Hubris theories are also considered to have an effective CG. For instance, while acknowledging the adverse effects of hubris on CG, scholars do not deny that, under certain circumstances, a small amount of hubris may be beneficial [1]. In this respect, other scholars stress the importance of transparency and reasonability of remunerations and nominations subject to CG [17,18].

Furthermore, scholars also underline that herding negatively affects CG, especially in less mature entities and less procedurally formalised CG, resulting in lower market values for such organisations [19]. Inevitably, managerial entrenchment risks must also be considered [20]. In this context, CG is highly susceptible to the risks associated with agency theory. In this respect, CG represents a set of functions to govern an organisation efficiently and effectively. Consequently, CG is intended to act as a mechanism to tackle agency problems. However, stewardship theory dictates that the agent (director) should prioritise the interests of the principal (shareholders) before the interests of other stakeholders [21]. Therefore, not duly formalised CG practices might expose the risk of compliance with the expectations of the employees, shareholders, other stakeholders, and society [17,19,21]. Most of the referenced scholars in this article also concede these findings when exploring the context of CG in pursuit of the improved performance of undertakings. However, to comprehensively understand CG, it is essential to know its key elements, with a pivotal focus on the board structure, which is subject to supervision, control, and decision-making levels. Respectively, different types of board structures are used worldwide. The two main types are multi-tier and unitary boards (Figure 1).

The United States mainly utilises the unitary structure with a single board of directors [22]. In contrast, the European Union (EU) frequently uses the multi-tier structure, which divides the board of directors into two bodies: the supervisory and management boards [23]. Each type has its advantages. The supervisory board allows executive directors to participate in their meetings without limitations. However, this can lead to the necessity of holding additional meetings to reach decisions. As a result, organising all the required meetings and communication between the two boards becomes expensive. In the case of the unitary board structure, the organisation has only one board, which includes executive directors and individuals responsible for operational duties within the business. However, a one-tier board can cause a risk of the CEO overseeing more than one position within the board, resulting in the risk of mismanagement. Akstinate [1] states that a two-tier board reduces such a risk by splitting the supervisory and day-to-day management functions between the two bodies.

For decades, scholars have underscored the board’s principal responsibility in guaranteeing that enterprise management aligns with its shareholders’ best interests [24,25]. However, currently, the requirements of sustainable development have been observed to be more important than the interests of shareholders [26,27]. In this context, prioritising sustainable development over shareholder interests should define effective CG. The latter would also align with the contemporary societal theory, emphasising the obligation of any organisation to all the interested parties affected by its actions [28]. Compliance with shareholder requirements should only be subject to CG efficiency. To measure the latter, scholars examine board composition, leadership structure, member traits, and operational procedures [29,30].

Board composition typically includes executive directors, non-executive directors, the presence of women on the board, and the size of the board. The findings indicate that increasing the representation of women and non-executive members on larger boards may reduce business risk and improve the performance of the control function [31,32,33]. This highlights the potential advantages of promoting diversity and inclusivity within the leadership of an organisation. Boards with higher non-executive/executive director ratios, more women, and larger sizes tend to be more effective and productive.

Regarding board leadership structure, scholars consider the CEO’s competence and level of authority, block ownership, and board executive ownership. Past studies have shown that companies with higher board ownership have lower earnings quality and higher firm risk [3,34]. According to the theory of management entrenchment, when the percentage of equity held by institutional investors exceeds the mean value, the risk level of business operations increases [4]. In this regard, scholars also note that enterprises are likely to manipulate information disclosure to the public concerning business risks and the dynamics of competition within their markets [35,36]. The latter insights are also valuable in the context of agency theory [17,19,21], whereas CG must be committed to developing businesses competitively and sustainably.

The third attribute encompasses variables associated with board members’ age and length of service. Higher age and longer proxy careers are linked with reduced business risk [2,34].

The last component of the board process encompasses variables associated with board meetings, specifically the attendance and frequency of audit committee gatherings. Conducting advisory meetings can effectively mitigate corporate risk [37].

Considering the above context, scientists argue that various CG evaluation models can be applied, adapted, and even partially replaced by others to evaluate the board’s performance and identify areas for improvement [5,38]. In this respect, the coexistence of multiple governance models implies that employing a specific one is not practical or effective. As a result, scientists use a variety of criteria, such as meetings, duality, and the inclusion of non-executive members, to calculate governance scores [39]. However, each research work into governance is distinct, as it focuses on a particular aspect and does not reflect any specific players of any economic sector. For instance, studies [27,40] specifically looked at the audit committee and the frequency of their meetings as a component. Others focus on diversity and ownership [31] or the impact of CG on the correlation between financial performance and capital structure [41]. Crifo et al. [42] emphasise the board of directors and investor relations officers, revealing challenges about sustainability and the required leadership efforts. Aspects of public control concerning CG in state-owned organisations are also studied [43]. In this respect, Xiong et al. [44] developed CG and market power indicators and examined the mechanism of costs of job loss. Other scholars [45] focus on the dynamics of ownership structure in state-owned organisations and corporate environmental responsibility (CER) and indicate that a larger separation of ownership impedes CER in state-owned enterprises (SOEs). In this regard, no specific focus was on the energy sector or related organisations.

To explore the CG of entities in India and Gulf Corporation Council (GCC) countries [46], the scholars aimed to develop a tool to identify efficient businesses that implement the best governance features, which could serve as a benchmark for other companies. The proposed solutions have led to the dualistic description of two opposing governance systems. The first system is characterised by the CEO’s control over the board and concentration of ownership. In contrast, the second system is defined by inside control efficiency. Gupta and Pandey [39] focused on assessing the impact of CG benchmarks on working capital management in publicly traded enterprises.

Other scholars [47,48] integrated studies on the correlation of endogenous factors in CG, enterprise performance, and leverage. Their results showed that enterprise performance has a positive and substantial relationship with CG. Consequently, the market value of enterprise performance and the market-to-book ratio can be the primary moderators of the relationship between external governance mechanisms and, for instance, provisions concerning mergers and acquisitions [49]. The latter scholars also noted that greater intellectual depth and knowledge are linked to larger boards, which enhances performance and aids in decision-making. Conversely, the findings suggest no correlation between CG indicators and profitability or return on equity. The findings also imply no correlation between the CEO and board chairperson positions being held simultaneously and any aspect of enterprise performance.

Consequently, in the above context, performance measurement aims to support the management and implementation processes of corporate strategy, which CG oversees. To ensure efficient and effective performance metrics for CG, indicators must align with the corporate strategy’s strategic, tactical and operational objectives. According to scholars [50], the right metrics of key performance indicators (KPIs) should help uncover and understand the drawbacks of CG performance, identify specific performance concerns, compare actual conditions to goals, and indicate actions to be resolved. Therefore, the continuous monitoring of performance measurement variables is imperative for timely responsiveness, necessitating around-the-clock attention. Consequently, the metrics should indicate how well the team (or an individual) is doing in a specific area, what it needs to do, what exactly it has achieved, and what it needs to do to improve performance [51]. The CG’s timely reaction to these latter aspects will impact its effectiveness. Considering the above context, CG inevitably needs a strategic performance measurement system that converts the strategy into metrics, targets, and activities, focusing on alignment, leadership, and the organisation’s purpose. For instance, in this regard, the scholars [52] integrated a compilation of sustainable improvement standards and a collection of overall performance indicators based on the balanced scorecard approach, of which scholars have explored several generations [53]. Others focus on the dynamics of CG and the cost efficiency of banking [54].

Following the above context, the author points out that there has been little emphasis by scholarly research on state-owned power suppliers (as in the focus of this paper) or the energy sector (if considering the broader scope) regarding CG performance assessment concerning strategic, tactical, and operational business results. Specifically, only 102 papers were found on Web of Science and 51 in the Scopus database (as of 4 July 2024) relating to state-owned organisations or the energy sector. The total number of relevant papers is likely even lower due to potential overlap between the databases. Additionally, only a few of these papers directly addressed the topic. For instance, in the case of Poland, scholars [55] focused on institutional investors, board sizes, and state ownership and investigated their impact on capital structure concerning debt levels, resulting in the need for further research. Others examined the influence of independent supervisory boards on transformations in the energy sector worldwide [56] and concluded that independent supervisory boards only direct organisations through the changing regulatory environment, with other findings being equivocal due to regulatory specifics. In this respect, Shahbaz et al. [57] found a positive influence of independent board members and female directors on CG performance in the energy sector worldwide and from an environmental and social responsibility perspective. Makridou et al. [58], considering environmental, social and governance (ESG) factors and corporate financial performance in the energy sector, also indicate a positive but not significant association of CG on the economic performance of the organisations.

Nevertheless, in conclusion, the scholarly research addressed in this subsection also makes it apparent that CG and business performance variables are prevalent across different economic sectors. However, no concrete studies were found that were subject to the impact of CG variables on the KPIs of state-owned entities acting in the energy sector. Also, the reviewed scholarly papers reveal that while striving to investigate CG variables in complex settings (or specific groups or numbers of variables), the authors, in principle, result in making valuable conclusions only concerning one or few CG variables. The individual studies struggled to conclude that all CG variables investigated have positive, negative, or minor or no impacts in the holistic context or conclude the need for further research.

Based on these arguments above, hypotheses were formulated focusing on concrete CG variables. In this regard, the following hypotheses were derived to be tested in the subsequent sections:

Hypothesis 1 (H1).

Larger and more diverse boards positively impact the performance of state-owned power suppliers and increase their net income.

Hypothesis 2 (H2).

The presence of independent members contributes to the net income growth of state-owned power suppliers.

Hypothesis 3 (H3).

The frequency of audit meetings positively impacts the earnings but does not necessarily increase the earnings of state-owned power suppliers.

Hypothesis 4 (H4).

A larger audit committee can contribute to CG’s decision-making process.

Consequently, the respective models will be constructed in the next section to investigate the CG variables concerning the KPIs of the selected energy producers. Additionally, specific indicators for gauging the performance of energy sector enterprises must be discerned. These aspects shall be addressed in the following subsection.

2.2. The Specifics of Performance Measurement in the Power Sector

Per Ghiasi et al. [59], the fundamental objective of a power system is to supply electricity to consumers while maintaining stringent standards of quality and uninterrupted service. The quality of service offered to consumers is directly impacted by the reliability of the power supply. The adoption of five energy packages in the EU, subject to the liberalisation of the energy market [10] over the past three decades, has brought about significant changes in the electricity market, resulting in a transformed profile. This has resulted in changes in electricity consumption, losses, peak load, energy exchange balance, electricity generation, voltage profile, and customer base. The specifics of consumption dynamics are also affected [60], and the latter must be considered for sustainable and competitive power supply. To ensure high-quality services at a reasonable cost, utilities require an effective network operations management system that enables the timely identification of problem areas and the implementation of targeted corrective actions to ensure a reliable power supply. The reliability of the power supply system is contingent upon the effective coordination of its three integral components: power generation, transmission, and distribution. In this respect, scholars [59,61] underscore the importance of reliability parameters, specifically (i) System Average Interruption Duration Index (SAIDI), (ii) System Average Interruption Frequency Index (SAIFI), and (iii) Customer Average Interruption Duration Index (CAIDI). The latter indices evaluate the effectiveness of the electrical power network and should be computed at the operational district levels. The calculations must be carried out continuously to ensure the reliability of power supply indices for use at strategic, tactical, and operational levels in terms of the planning and execution of the tasks. According to scholars [61,62], monitoring and benchmarking reliability performance must be a standard procedure in the electric utility sector. Consequently, SAIDI (the ratio of total customers served to the duration of customer interruptions) can be calculated in the following manner (Equation (1)):

SAIDI = U_T/N_T,

(1)

where U_i is the total annual outage time per location and N_T is the total number of customers served per location.

The SAIDI (Equation (1)) denotes the average duration of customer disruptions per year, typically measured in terms of customer minutes or customer hours of disruption. A reduction in the frequency or duration of interruptions can improve SAIDI. Accordingly, the SAIDI threshold may ascertain the aggregate number of calendar days throughout which the system’s design or operational limits are surpassed [63]. The indicator shows the dependability of the power supply network and reveals the amount of energy not supplied. Regarding strategic management of the power industry enterprises, the external and internal factors affecting the changes in SAIDI must be addressed. Consequently, the System Average Interruption Frequency Index (SAIFI) must be considered, namely (Equation (2))

SAIFI = F_T/N_T,

(2)

where F_T is the total number of failures in the location and N_T is the total number of customers served per location.

The SAIFI (Equation (2)) indicates the frequency with which a customer encounters sustained service interruptions within a defined timeframe, typically spanning one year. There is flexibility in the definition of the location due to its ability to adapt to changes in the number of customers and the interruptions they experience. A feeder’s SAIFI, for instance, provides insight into the average number of disruptions experienced by a customer served by that feeder over a year. Furthermore, SAIFI also indicates the total number of customers in the service area enclosed by a substation or distribution system. As a result, the power suppliers must minimise the restoration time for the electricity supply. In this regard, the Customer Average Interruption Duration Index (CAIDI), the ratio of SAIDI to SAIFI [61,63], must be considered. CAIDI represents the average duration required to restore customers’ service following an extended outage. To enhance CAIDI, it is necessary to minimise the duration of disruptions and expedite crew response. This necessitates improving planning and the organisation of individual tasks per crew member, including potential interactions with other on-site teams.

Following the context above, a meticulous focus on operational KPIs is consistent with the fundamental objective of power supply entities: to deliver electricity in a stable, dependable, and standardised manner. Subsequently, CG must be pivotal in directing the strategies that directly influence these KPIs. In this regard, prioritising SAIDI, SAIFI, and CAIDI in CG frameworks is not merely a technical necessity but a strategic imperative, ensuring power suppliers navigate challenges, adapt to market dynamics, and cost-effectively deliver unparalleled service quality. Therefore, these operational KPIs are essential for enhancing CG, guiding the entities towards sustained effective performance and prioritising stakeholder satisfaction. However, when dealing with competition from the liberalisation of the energy market [10], power suppliers must focus on meeting the customers’ needs and preferences to retain existing customers and launch new businesses (e.g., providing both fossil fuel-based and renewable energy-based solutions). Enterprises currently employ approaches such as monitoring consumers’ online behaviour and implementing voluntary feedback programmes to align with customer requirements, as noted by scholars [64]. Nevertheless, structured surveys were also used, enabling entities to question targeted topics and customers to share their experiences with the enterprise. Based on the gathered representative customer data, businesses have the potential to elevate the satisfaction levels of their clientele with their products and service offerings. The Net Promoter Score (NPS), a survey-based metric, can effectively forecast future sales growth [64,65]. The measurement of NPS involves a straightforward process, relying on responses to a single inquiry. Customers are requested to provide ratings on a scale from 0 to 10, with each rating falling into three categories (i.e., into promoters, passives, or detractors, accordingly). Then, NPS is calculated by subtracting the number of “detractors” from the number of “promoters” and then dividing by the total sample size, as shown in Equation (3), namely

NPS = Pr_T − Dtr_T/Sample size,

(3)

where Pr_T is the total number of promoters (respondents providing ratings of 9 or 10) and Dtr_T is the total number of detractors (respondents providing ratings below 6).

Consequently, NPS can be used to determine customer loyalty and improve entity performance [66]. In this regard, managers should improve NPS, aiming at future sales growth [64,66]. Scholars also debate whether structured surveys (such as those focusing on deeper client satisfaction analysis) could be used together with NPS [67,68]. Also, the study by Agag et al. [69] reveals the reliability of using NPS and emphasises utilising a combination of varied customer feedback metrics. As a result, enhancing the possibility of providing recommendations may lead to actual suggestions, which can positively influence other potential customers and ultimately result in new sales and purchases [70]. In this context, the focus should be on current customers to improve NPS and those familiar with the goods and services. Hence, all potential customers must be considered, according to Baehre [71].

Considering the insights presented in this subsection, it is evident that the assessment of power suppliers’ operational efficiency hinges significantly on metrics such as SAIDI, SAIFI, and NPS. These metrics directly influence the enterprises’ strategic objectives and are, therefore, considered to fall within the purview of corporate governance responsibility. Consequently, the following hypotheses are formulated in addition to those stated in Section 2.1, to be addressed in the next sections:

Hypothesis 5 (H5).

CG positively affects SAIDI, SAIFI, and NPS. Improving these indicators should contribute to the business performance of state-owned power suppliers.

Hypothesis 6 (H6).

The presence of independent board members and greater board gender diversity improve SAIDI, SAIFI, and NPS.

The following section will also consider SAIDI, SAIFI, and NPS when developing models to explore specific energy producers’ CG variables and KPIs.

3. Materials and Methods

Scholarly research concerning state-owned power supplies is fragmented and scarce. In our study, five power suppliers, Lithuanian Ignitis Group, Latvian Latvenergo, Estonian Eesti Energia, Danish Orsted, and Swedish Vattenfall, were selected to examine the CG impact on enterprise performance variables. All the enterprises in question are leading enterprises in their respective homelands. The state wholly owns Latvenergo, Eesti Energia, and Vattenfall. The state is the majority shareholder in Orsted (51%) and Ignitis (75%).

Descriptive statistics, Pearson correlation, and regression analysis were employed to process the data for 2019–2023. Descriptive statistics allow for determining the data’s tendency, variability, and distribution. The Pearson correlation will enable us to establish and measure a linear relationship among variables. With a linear correlation established among the variables in question, multiple linear regression analysis is then carried out to identify cause–effect relationships among variables (

Table 1. Description and measurements of variables of the CG and enterprise performance.

Variables	Abbreviation	Measurements
Dependent Variables
Revenue	RE	Sales or turnover refers to the total amount of money earned by an enterprise from its primary activities.
Net profit	NI	Net income, or net earnings, refers to the amount of revenue a company has earned after deducting all expenses, including taxes and interest.
Earnings before interest, taxes, depreciation, and amortisation	EBITDA	The metric measures profitability of the enterprise before accounting for certain non-operating expenses, representing its operating performance.
Net debt/EBITDA	EBDE	The ratio measures an enterprise’s leverage by dividing its cash or cash equivalents by its EBITDA after subtracting interest-bearing liabilities. The ratio shows how many years it would take for an enterprise to pay off its debt, assuming net debt and EBITDA remain constant.
System Average Interruption Duration Index (SAIDI)	SDI	Metric for measuring the average duration of power outages per customer in an electrical distribution system.
System Average Interruption Frequency Index (SAIFI)	SFI	Metric that measures the average number of power outages a customer experiences in an electrical distribution network.
Net promoter score (NPS)	NPS	Market research metric that measures the likelihood of recommending a company, product, or service to others.
Independent Variables
Board size	BSI	Total number of board members
Executive members	EXE	Total number of executive directors on the board.
Independent members	NEXE	Members who are not affiliated with the company
Male directors	MDI	Total number of male board members
Female directors	FDI	Total number of female board members
Board gender diversity	BGD	Percentage of male and female members on the board
Board meeting frequency	BMF	The total number of board meetings held during the specified period.
Audit meeting frequency	AMF	The total number of audit meetings held during the relevant period.
Audit committee size	ACS	Number of audit committee members

).

Firstly, data from publicly available annual reports of selected studies [72,73,74,75,76] were used to generate 25 observations (5 companies × 5 years).

Secondly, the SAIDI, SAIFI, and NPS were chosen to measure the performance of enterprises in the power sector, focusing on financial, operational, and socio-economic aspects. Data on these variables are collected by conducting unstructured interviews with the relevant enterprises’ relevant officers. Interview questions and findings are presented in the next section.

Thirdly, descriptive statistics were produced by combining data from publicly available annual reports of selected enterprises and the interview findings on SAIDI, SAIFI, and NPS. As a result, the descriptive statistics will indicate the dynamics and trends in the data within the dataset.

Consequently, considering the findings of the previous section, CG’s influence on business performance will be examined based on the sets of variables presented in Table 1.

The CG variables (independent) are the board size, executive members, independent members, male directors, female directors, board gender diversity, board meeting frequency, audit meeting frequency, and audit committee size.

The enterprise performance variables (dependent) are revenue, net profit, EBITDA, net debt/EBITDA, SAIDI, SAIFI, and NPS.

Next, Pearson correlation and multiple linear regression analysis were carried out. To measure and assess the performance of enterprises quantitatively, understand trends and patterns over time, and identify possible regularities, the research utilised Ordinary Least Squares Regression to clarify the correlation between selected performance and CG variables. The following model was built,

Y = α + β_1×1 + β₂X₂ + β₃X₃ + β₄X₄ + β₅X₅ + β₆X₆ + β₇X₇ + β₈X₈ + β₉X₉ + ε,

(4)

where the variables in this model are as follows:

-

Y: an aggregated variable consisting of CG variables.

-

α: the value of Y when all independent values equal 0.

-

β₁ through β₉: quantified dependent values that each CG variable has on the KPIs.

-

X₁: the total count of board members.

-

X₂: the count of executive directors on the board.

-

X₃: the count of members not affiliated with the company.

-

X₄: the total count of males acting as board members.

-

X₅: the total count of females acting as board members.

-

X₆: the percentage of male and female members on the board.

-

X₇: the total count of board meetings held during the specified period.

-

X₈: the total count of audit meetings held during the relevant period.

-

X₉: the number of audit committee members.

-

ε: an error term.

The following multiple linear regression models can be used concerning respective KPIs:

RE = α + β₁BSI + β₂EXE + β₃NEXE + β₄MDI + β₅FDI + β₆BGD + β₇BMF + β₈AMF + β₉ACS + ε,

(5)

NI = α + β₁BSI + β₂EXE + β₃NEXE + β₄MDI + β₅FDI + β₆BGD + β₇BMF + β₈AMF + β₉ACS + ε,

(6)

EBIDTA = α + β₁BSI + β₂EXE + β₃NEXE + β₄MDI + β₅FDI + β₆BGD + β₇BMF + β₈AMF + β₉ACS + ε,

(7)

EBDE = α + β₁BSI + β₂EXE + β₃NEXE + β₄MDI + β₅FDI + β₆BGD + β₇BMF + β₈AMF + β₉ACS + ε,

(8)

SDI = α + β₁BSI + β₂EXE + β₃NEXE + β₄MDI + β₅FDI + β₆BGD + β₇BMF + β₈AMF + β₉ACS + ε,

(9)

SFI = α + β₁BSI + β₂EXE + β₃NEXE + β₄MDI + β₅FDI + β₆BGD + β₇BMF + β₈AMF + β₉ACS + ε,

(10)

NPS = α + β₁BSI + β₂EXE + β₃NEXE + β₄MDI + β₅FDI + β₆BGD + β₇BMF + β₈AMF + β₉ACS + ε,

(11)

Using the above models, all calculations, commencing with descriptive statistics, Pearson correlation analysis, and regression analysis, were executed using Python programming. The subsequent section details the experiment’s outcomes and discusses the findings.

4. Results

Data collected from five power enterprises were subject to descriptive statistics. The dataset included 25 observations and presents a comprehensive overview of the companies’ performance and CG.

Data concerning the SAIDI, SAIFI, and NPS were collected through phone interviews with the relevant officers of the respective divisions of the enterprises. At least one relevant officer (who the author was directed to by the officials of the enterprises investigated) was interviewed per enterprise. Officers were asked one question concerning each variable, namely

“What was the system average interruption duration per annum/quarter/month for the years 2019, 2020, 2021, 2022, and 2023?”, subject to SAIDI.

“What was the system average interruption frequency per annum/quarter/month for the years 2019, 2020, 2021, 2022, and 2023?”, subject to SAIFI.

“What was the net promoter score per annum/quarter/month for 2019, 2020, 2021, 2022, and 2023?” It was explained to officers that the latter index is based on asking the client only one question: “How likely is it that you would recommend services X of organisation Y to a friend or colleague?”

To facilitate cooperation in providing required data, the meaning of each variable was explained upon request to the officers. Even after initial conversations, further clarifications were made through additional questions, making the interviews unstructured in principle. The sole purpose of the interviews was to obtain the necessary data. Under a nondisclosure agreement, the author was not authorised to use the data to compare entities or disclose their names publicly. It is important to note that the data collected relied solely on the goodwill and discretion of the officers. For example, there was no intention to provide the monthly or quarterly data. The data collected are presented in

Table 2. Collected data on SAIDI, SAIFI, and NPS.

Enterprise and Years	Enterprise 1			Enterprise 2			Enterprise 3			Enterprise 4			Enterprise 5
	SAIDI	SAIFI	NPS	SAIDI	SAIFI	NPS	SAIDI	SAIFI	NPS	SAIDI	SAIFI	NPS	SAIDI	SAIFI	NPS
2019	405.82	7.50	1.00	223.00	3.80	11.00	234.00	0.63	14.05	39.00	0.57	22.00	27.00	0.42	39.00
2020	294.00	5.40	9.00	214.00	3.71	7.00	201.00	0.87	12.00	43.00	0.73	26.00	33.00	0.81	29.00
2021	346.99	7.30	6.00	231.00	1.78	13.00	238.00	0.59	17.00	47.00	0.59	24.00	35.00	0.51	36.02
2022	439.00	8.90	12.10	208.00	2.84	10.00	208.00	0.80	12.01	43.00	0.68	28.00	34.00	0.74	30.00
2023	384.00	5.70	7.00	234.00	0.91	14.00	236.00	0.59	15.00	41.00	0.54	24.00	35.00	0.49	35.00

Source: Author based on the results of the unstructured interview.

.

Data from the interviews were combined with data gathered from publicly available information on selected enterprises to create descriptive statistics (

Table 3. Descriptive statistics.

Variable	N	Min	Max	Mean	Std
Board	25	7	15	11.12	2.147091
Executive directors	25	2	8	5	1.554563
Independent members	25	3	8	5.04	1.540563
Male directors	25	5	11	8.2	1.825742
Female directors	25	0	6	2.92	1.93477
Board gender diversity	25	0	0.5	0.2536	0.154485
Supervisory meeting frequency	25	10	30	13.96	4.586938
Audit meeting frequency	25	4	24	9.44	5.385784
Audit committee size	25	3	6	4.6	0.763763
Revenue	25	774.1	21,089	6117.747	6492.578
Net income	25	−22	4225.144	711.913	1047.013
EBITDA	25	149.9	6669.52	1619.897	1833.359
Net debt/EBITDA	25	0.39	7.31	2.9544	1.639345
SAIDI	25	27	439	178.9536	105.0683
SAIFI	25	0.42	8.9	2.2076	1.774374
NPS	25	1	39	18.1272	11.00576

Source: Author.

).

According to the findings, board size averages around 11 members, with variable gender composition requiring nuanced leadership. The consistent yet adaptive governance structure is reflected in audit committees of three to six members. As to the data presented in Table 3, the revenue figures show significant variation, with values ranging from 774.10 million to 21,089 million. This indicates that the enterprises have different market positions and operate in diverse industries. The net income, ranging from negative 22 to 4225.144 million, shows fluctuations in profitability. EBITDA ranges from 149.90 to 6669.520 million, showcasing diverse operational efficiency. The ratio of net debt to EBITDA, which indicates the level of leverage, ranges from 0.39 to 7.310.

Accordingly, indices measuring the reliability of power supply, such as SAIDI and SAIFI, exhibit considerable variability, reflecting the challenges of strategies employed by the power suppliers investigated when ensuring service reliability. In this regard, NPS, for example, ranges from 1 to 39, highlighting diverse approaches to stakeholder management.

Consequently, a Pearson correlation analysis was conducted to explore the key variables related to CG, financial performance, and governance efficiency indices. The resulting Pearson correlation coefficients (

Table 4. Pearson correlation analysis.

(a) R-Square Values
	BSI	EXE	NEXE	MDI	FDI	BGD	BMF	AMF	ACS	RE	NI	EBITDA	EBDE	SDI	SFI	NPS
BSI	1.000
EXE	0.811	1.000
NEXE	0.288	0.226	1.000
MDI	0.536	0.587	−0.225	1.000
FDI	0.604	0.346	0.532	−0.349	1.000
BGD	0.347	0.069	0.463	−0.598	0.949	1.000
BMF	−0.139	−0.257	−0.076	−0.312	0.140	0.206	1.000
AMF	0.105	−0.059	−0.447	−0.150	0.259	0.300	0.566	1.000
ACS	0.195	0.533	−0.020	0.056	0.164	0.036	0.034	0.186	1.000
RE	0.471	0.434	0.739	−0.022	0.544	0.459	−0.341	−0.479	−0.169	1.000
NI	0.265	0.107	0.684	−0.130	0.417	0.371	−0.203	−0.413	−0.223	0.602	1.000
EBITDA	0.457	0.359	0.798	−0.040	0.546	0.456	−0.335	−0.509	−0.156	0.883	0.873	1.000
EBDE	−0.122	−0.175	−0.168	−0.188	0.042	0.111	0.074	0.185	−0.386	−0.043	−0.078	−0.120	1.000
SDI	0.242	0.399	−0.450	0.643	−0.338	−0.491	−0.094	−0.002	0.343	−0.310	−0.440	−0.383	−0.174	1.000
SFI	0.323	0.444	−0.092	0.487	−0.101	−0.256	−0.278	−0.177	0.253	0.038	−0.218	0.033	−0.057	0.316	1.000
NPS	−0.127	−0.437	−0.298	0.113	−0.248	−0.167	−0.018	0.047	−0.760	−0.130	−0.017	−0.125	0.299	−0.106	−0.249	1.000
(b) p-values
	BSI	EXE	NEXE	MDI	FDI	BGD	BMF	AMF	ACS	RE	NI	EBITDA	EBDE	SDI	SFI	NPS
BSI	0.000
EXE	0.000	0.000
NEXE	0.162	0.277	0.000
MDI	0.006	0.002	0.279	0.000
FDI	0.001	0.090	0.006	0.087	0.000
BGD	0.090	0.742	0.020	0.002	0.000	0.000
BMF	0.507	0.215	0.717	0.128	0.503	0.323	0.000
AMF	0.616	0.778	0.025	0.474	0.212	0.145	0.003	0.000
ACS	0.351	0.006	0.925	0.791	0.435	0.865	0.870	0.374	0.000
RE	0.017	0.030	0.000	0.915	0.005	0.021	0.095	0.015	0.421	0.000
NI	0.201	0.611	0.000	0.535	0.038	0.068	0.330	0.040	0.283	0.001	0.000
EBITDA	0.022	0.078	0.000	0.849	0.005	0.022	0.102	0.009	0.455	0.000	0.000	0.000
EBDE	0.560	0.403	0.423	0.367	0.842	0.596	0.726	0.376	0.056	0.840	0.712	0.566	0.000
SDI	0.244	0.048	0.024	0.001	0.098	0.013	0.653	0.993	0.093	0.132	0.028	0.059	0.406	0.000
SFI	0.115	0.026	0.663	0.013	0.631	0.216	0.179	0.396	0.223	0.858	0.296	0.876	0.785	0.124	0.000
NPS	0.545	0.029	0.147	0.590	0.232	0.425	0.932	0.822	0.000	0.534	0.935	0.553	0.147	0.613	0.231	0.000

Source: Author.

a (indicates the strength and direction of associations) and 4b (specifies the corresponding p-values subject to the statistical significance of the relationships)) provide several insights concerning the interactions of dependent and independent variables in the context of CG.

After considering the dependent variables (Table 4a,b), the following was observed:

• The revenue variable (RE) had a statistically substantial positive correlation with board size (r = 0.471, p = 0.017), executive members (r = 0.434, p = 0.030), independent members (r = 0.739, p = 0.000), female directors (r = 0.544, p = 0.005), and the board gender diversity variable (r = 0.459, p = 0.021). The latter can be considered as confirming H1. A statistically significant negative correlation has been identified between RE and the frequency of audit meetings (r = −0.479, p = 0.015), which confirms H3.
• Also, there is a significant positive correlation between net income (NI) and independent members (r = 0.684, p = 0.000), female directors (r = 0.417, p = 0.038), and RE (r = 0.602, p = 0.001). The latter, in principle, confirms H2. On the other hand, there is a negative association between NI and the audit meeting frequency (r = −0.413, p = 0.040), which confirms H3.
• In terms of EBITDA, there are substantial positive correlations with the variables of independent members (r = 0.798, p = 0.000) and female directors (r = 0.546, p = 0.005), board gender diversity (r = 0.456, p = 0.022) and board size (r = 0.457, p = 0.022), and RE (r = 0.883, p = 0.000) and NI (r = 0.873, p = 0.000). The latter confirms H2. No significant relationship was observed between the size of the audit committee and EBITDA (correlation coefficient r = −0.156, p-value = 0.455). This does not support H4. However, there is a negative association between EBITDA and the frequency of audit meetings (correlation coefficient r = −0.509, p-value = 0.009), hence confirming H3.
• Net debt/EBITDA (EBDE) negatively correlated with audit committee size (r = −0.386, p = 0.056). The latter can be considered as confirming H4.
• SAIDI exhibited a significant positive correlation with male directors (r = 0.643, p = 0.001) and a smaller one with executive members (r = 0.399, p = 0.048). The latter findings can be considered to support H6. Also, a negative correlation was found with independent members (r = −0.450, p = 0.024) and the board gender diversity variable (r = −0.491, p = 0.01), NI (r = −0.440, p = 0.028), and EBITDA (r = −0.383, p = 0.059).
• SAIFI positively correlated with executive members (r = 0.444, p = 0.029) and male directors (r = 0.487, p = 0.013), partially supporting H6.
• NPS displayed significant negative correlations with audit committee size (r = −0.760, p = 0.000) and executive members (r = −0.437, p = 0.029). The latter findings do not support H5 and partially reveal no support for H4. In this respect, the need for a greater formalisation of CG could be considered, improving accountability per concrete objectives to be achieved.

Analysing results (Table 4a,b) for independent variables, statistically significant positive associations of board size (BSI) were observed with executive members (r = 0.811, p = 0.000) and male (r = 0.536, p = 0.006) and female (r = 0.604, p = 0.001) directors. Also, significant positive correlations of executive members (EXE) were found with male directors (r = 0.587, p = 0.002) and audit committee size (r = 0.533, p = 0.006). The latter findings can be considered to ensure the reasonable balance of diversification of competencies required for effective CG performance.

The results above show significant correlations. For instance, the revenue variable is positively associated with board size, executives, independent members of the board, female directors, and gender diversity variables.

This suggests that larger and more diverse boards should positively impact the performance of investigated power suppliers (supporting H1). As a result, such entities tend to outperform those with smaller and less diverse boards. The latter statement is further reinforced by the notable positive correlations between EBITDA and board size, independent members, female directors, and gender diversity. These variables may positively affect the earnings of the enterprises. Respectively, H2 can also be considered confirmed. However, it should be noted that the frequency of audit meetings should not be regarded as a factor that enhances earnings due to the observed negative correlation with EBITDA (supporting H3). Concerning the size of the audit committee, certain support was found for H4 in terms of EBDE.

Regarding SAIDI and SAIFI, the author considers the observed positive correlation between male directors and executive members from a managerial point of view. Specifically, involving male directors with engineering backgrounds in decision-making should minimise the duration and the number of power outages. In this respect, H6 can be supported. Furthermore, the latter insight can be supported by a negative correlation found between SAIDI and EBITDA (this also partially confirms H5), the independent board members, and the board gender diversity (and no correlation in this regard to consider with SAIFI).

In the case of NPS, a negative correlation with audit committee size may be considered in terms of the level of control and formalisation of the processes. The latter may hamper the flexibility of the entities’ operations and, as a result, negatively affect customer satisfaction. In this respect, there is no support for H5, and no support in terms of NPS can be indicated for H4.

Nevertheless, given the context above, it is imperative to exercise caution when interpreting causation from correlation. It is crucial to consider contextual factors to establish causative relationships and understand the implications for CG. Thus, regression analysis was performed next.

All the KPI-related models presented in the previous section underwent regression analysis. The regression analysis showed statistically significant correlations between CG variables and revenue, EBITDA, SAIDI, and NPS ratios, with R-squared values above 50%. However, due to the high collinearity between the board size and the number of male and female directors, the latter were removed from the models, leaving seven independent variables. Therefore, the following multiple linear regression models were examined:

RE = α + β₁BSI + β₂EXE + β₃NEXE + β₆BGD + β₇BMF + β₈AMF + β₉ACS + ε,

(12)

NI = α + β₁BSI + β₂EXE + β₃NEXE + β₆BGD + β₇BMF + β₈AMF + β₉ACS + ε,

(13)

EBIDTA = α + β₁BSI + β₂EXE + β₃NEXE + β₆BGD + β₇BMF + β₈AMF + β₉ACS + ε,

(14)

EBIDTA = α + β₁BSI + β₂EXE + β₃NEXE + β₆BGD + β₇BMF + β₈AMF + β₉ACS + ε,

(15)

SDI = α + β₁BSI + β₂EXE + β₃NEXE + β₆BGD + β₇BMF + β₈AMF + β₉ACS + ε,

(16)

SFI = α + β₁BSI + β₂EXE + β₃NEXE + β₆BGD + β₇BMF + β₈AMF + β₉ACS + ε,

(17)

NPS = α + β₁BSI + β₂EXE + β₃NEXE + β₆BGD + β₇BMF + β₈AMF + β₉ACS + ε,

(18)

Regarding the revenue (RE) model (

Table 5. Revenue regression model (RE model).

Variable	Coefficient	Std_Error	t-Value	p-Value	95%_CI_Lower	95%_CI_Upper
const	2429.1869	1.49 × 104	0.164	0.872	−2.91 × 104	3.39 × 104
BSI	−119.5853	973.418	−0.123	0.904	−2183.139	1943.968
EXE	2185.0964	688.029	3.176	0.006	726.539	3643.653
NEXE	1111.5549	610.312	1.821	0.087	−182.249	2405.358
BGD	3.23 × 104	4.07 × 104	0.795	0.438	−5.39 × 104	1.19 × 105
BMF	−75.6545	140.463	−0.539	0.598	−373.423	222.114
AMF	−464.649	154.278	−3.012	0.008	−791.704	−137.594
ACS	−2535.7573	656.277	−3.864	0.001	−3927.003	−1144.512
Omnibus	Durbin_Watson	JB	Skew	Prob_JB	Kurtosis	Cond_No
1.562	1.951	1.257	−0.526	0.533	2.684	2.21 × 103

Source: Author.

), the positive correlation between executive members (EXE) and revenue (RE) indicates a significant impact.

However, other variables, such as audit committee size (ACS) and audit meeting frequency (AMF), show negative correlations.

A significant positive correlation with EXE suggests that a well-structured executive team positively influences revenue generation. This can be considered as partially supporting H1. Power entities reliant on strategic decision-making for revenue growth should prioritise the effectiveness of their executive board. A strong executive team can drive decision-making crucial for sustainable and diverse development in the power sector, where revenue is often tied to strategic, long-term, and capital-intensive investments and market dynamics.

In the net income (NI) model (

Table 6. Net income regression model (NI model).

Variable	Coefficient	Std_Error	t-Value	p-Value	95%_CI_Lower	95%_CI_Upper
const	−6158.1816	5392.248	−1.142	0.27	−1.76 × 104	5272.873
BSI	478.5638	353.331	1.354	0.194	−270.464	1227.592
EXE	−296.1077	249.741	−1.186	0.253	−825.534	233.319
NEXE	462.4796	221.531	2.088	0.053	−7.145	932.104
BGD	1.07 × 104	1.70 × 104	0.726	0.478	−2.06 × 104	4.20 × 104
BMF	−20.5391	50.985	−0.403	0.692	−128.623	87.545
AMF	−41.322	56	−0.738	0.471	−160.036	77.392
ACS	366.0173	238.215	1.536	0.144	−138.976	871.011
Omnibus	Durbin_Watson	JB	Skew	Prob_JB	Kurtosis	Cond_No
15.4	2.492	16.8	1.437	0.000225	5.804	2.21 × 103

Source: Author.

), a positive correlation with independent members (NEXE) can be considered. The latter can be an essential factor in avoiding conflicts of interest, improving the transparency of decision-making and reasonability of decisions made, complying with the interests of all stakeholders. The latter is of pivotal focus for state-owned power suppliers in compliance with corporate social responsibility and sustainability requirements. Moreover, the independent members can also contribute efficiently to achieving the strategic objectives regarding net income. In this respect, greater scrutiny can be expected concerning the management of fixed and variable costs for business activities. Consequently, the latter findings confirm H2.

Other variables do not indicate any correlation.

In the EBITDA model (

Table 7. EBITDA regression model (EBITDA model).

Variable	Coefficient	Std_Error	t-Value	p-Value	95%_CI_Lower	95%_CI_Upper
const	−1.05 × 104	6194.378	−1.698	0.109	−2.36 × 104	2616.076
BSI	744.1643	405.891	1.833	0.085	−116.286	1604.615
EXE	116.4893	286.891	0.406	0.69	−491.693	724.671
NEXE	711.6985	254.485	2.797	0.013	172.215	1251.182
BGD	2.64 × 104	1.70 × 104	1.557	0.139	−9541.211	6.23 × 104
BMF	−26.2423	58.57	−0.448	0.66	−150.404	97.92
AMF	−123.3628	64.33	−1.918	0.073	−259.737	13.011
ACS	239.5258	273.651	0.875	0.394	−340.589	819.64
Omnibus	Durbin_Watson	JB	Skew	Prob_JB	Kurtosis	Cond_No
2.502	1.967	1.106	0.383	0.575	3.689	2.21 × 103

Source: Author.

), a positive correlation between audit meeting frequency (AMF), NEXE (and EXE partially), and EBITDA could be considered. The latter can be regarded as not supporting H3, especially concerning NEXE, meaning that AMF contributes to the earnings increase of the investigated state-owned power suppliers. This suggests that audit meetings, independent members, and the executive team often contribute to achieving higher EBITDA. It signifies effective operational management and strategic decision-making. Concerning state-owned power suppliers, the latter findings contribute to ensuring the reasonability of the use of public funds. Also, as per the findings of the NI model (Table 5), stakeholders could expect more rationality in managing fixed and variable costs regarding business operations.

However, the other variables in the model exhibit weaker correlations.

The findings from the net debt/EBITDA (EBDE) model in

Table 8. EBITDA/Net debt regression model (EBDE model).

Variable	Coefficient	Std_Error	t-Value	p-Value	95%_CI_Lower	95%_CI_Upper
const	22.9378	11.165	2.054	0.057	−0.731	46.606
BSI	−1.2283	0.732	−1.679	0.113	−2.779	0.323
EXE	−0.1338	0.517	−0.259	0.799	−1.23	0.962
NEXE	−0.4955	0.459	−1.08	0.296	−1.468	0.477
BGD	−42.0674	30.561	−1.376	0.188	−106.855	22.72
BMF	−0.0974	0.106	−0.923	0.37	−0.321	0.126
AMF	0.1023	0.116	0.883	0.391	−0.143	0.348
ACS	−1.0791	0.493	−2.188	0.044	−2.125	−0.033
Omnibus	Durbin_Watson	JB	Skew	Prob_JB	Kurtosis	Cond_No
0.832	1.32	0.356	0.292	0.837	3.02	2.21 × 103

Source: Author.

have implications for effective debt management and financial stability. The results reveal a negative correlation between audit committee size (ACS) and EBDE, indicating that having a larger audit committee can lead to better governance efficiency in financial decision-making for the state-owned power suppliers in question. This can result in improved debt management and overall economic stability. The latter is especially critical considering the strict monetary policies prevailing in the EU, consequently raising investment costs and negatively affecting the purchasing power of customers. In this respect, the latter findings confirm H4.

Other variables show weaker correlations.

The correlations show that the CG variables have little effect on the SAIDI model (

Table 9. SAIDI regression model (SDI model).

Variable	Coefficient	Std_Error	t-Value	p-Value	95%_CI_Lower	95%_CI_Upper
const	74.9054	523.273	0.143	0.888	−1034.385	1184.195
BSI	18.5959	34.288	0.542	0.595	−54.091	91.283
EXE	14.4878	24.235	0.598	0.558	−36.889	65.864
NEXE	−42.0159	21.498	−1.954	0.068	−87.589	3.557
BGD	−251.9324	1432.344	−0.176	0.863	−3288.366	2784.501
BMF	5.6761	4.948	1.147	0.268	−4.813	16.165
AMF	−7.586	5.434	−1.396	0.182	−19.106	3.934
ACS	18.2284	23.117	0.789	0.442	−30.777	67.234
Omnibus	Durbin_Watson	JB	Skew	Prob_JB	Kurtosis	Cond_No
14.975	2.201	20.609	1.162	3.35 × 10−5	6.793	2.21 × 103

Source: Author.

). In this respect, no support was found for H5 or H6. Reliable service delivery is crucial for power generation and supply entities, necessitating effective governance. Therefore, specific CG responsibilities in this regard must be determined. Also, targeted research on the particular departments responsible for power supply operations is needed to gain further insights.

The SAIFI model presented in

Table 10. SAIFI regression model (SFI model).

Variable	Coefficient	Std_Error	t-Value	p-Value	95%_CI_Lower	95%_CI_Upper
const	4.7201	11.919	0.396	0.697	−20.548	29.988
BSI	−0.042	0.781	−0.054	0.958	−1.698	1.614
EXE	0.1365	0.552	0.247	0.808	−1.034	1.307
NEXE	−0.3743	0.49	−0.764	0.456	−1.412	0.664
BGD	−15.349	32.627	−0.47	0.644	−84.514	53.816
BMF	−0.0264	0.113	−0.234	0.818	−0.265	0.212
AMF	−0.0754	0.124	−0.609	0.551	−0.338	0.187
ACS	0.0918	0.527	0.174	0.864	−1.024	1.208
Omnibus	Durbin_Watson	JB	Skew	Prob_JB	Kurtosis	Cond_No
28.549	2.621	56.56	2.239	5.23 × 10−13	8.851	2.21 × 103

Source: Author.

did not show any significant correlations. This implies that the governance variables studied might not influence SAIFI to a relevant extent. In this respect, no support was found for H5 or H6. The lack of correlations indicates the need to further explore other and industry-specific governance factors to improve governance efficiency and reduce power system disruptions. As in the case of SAIDI, specific CG responsibilities must be defined for accountability and, hence, performance assessment, respecting the interests of all stakeholders.

The NPS model (

Table 11. NPS regression model (NPS model).

Variable	Coefficient	Std_Error	t-Value	p-Value	95%_CI_Lower	95%_CI_Upper
const	28.1384	57.608	0.488	0.632	−93.985	150.262
BSI	5.0156	3.775	1.329	0.203	−2.987	13.018
EXE	−7.0829	2.668	−2.655	0.017	−12.739	−1.427
NEXE	−1.8964	2.367	−0.801	0.435	−6.914	3.121
BGD	6.5071	157.689	0.041	0.968	−327.778	340.792
BMF	−0.0983	0.545	−0.181	0.859	−1.253	1.056
AMF	−0.2389	0.598	−0.399	0.695	−1.507	1.029
ACS	−2.9111	2.545	−1.144	0.269	−8.306	2.484
Omnibus	Durbin_Watson	JB	Skew	Prob_JB	Kurtosis	Cond_No
14.296	2.283	15.547	1.306	0.000421	5.846	2.21 × 103

Source: Author.

) shows a negative correlation between the presence of executives (EXE) and NPS, implying that the more executives there are, the lower the customer satisfaction levels tend to be. Also, the sufficiency of relevant qualifications of executives is to be considered. In this respect, no support was found for H5.

This raises concerns about the effectiveness of executive decision-making in promoting customer-centric strategies. To enhance the efficiency of governance in customer relations, it may be necessary to reconsider executive involvement in decision-making processes. In the energy sector, especially when the leading state-owned power suppliers are the focus, customer satisfaction can significantly impact public perception and regulatory relationships. Thus, CG should prioritise customer-centric decision-making to comply with stakeholders’ expectations.

5. Discussion

Considering all the context of the results presented in the previous section, as to descriptive statistics, it is evident that CG variables are diverse and that financial metrics exhibit significant variability, reflecting the complexity of the energy industry.

Consequently, according to Pearson correlation analysis, large and diverse boards contribute to better financial performance. The findings also indicate that audit meetings should be relatively regular, and their regularity should reflect the needs. Regarding mixed correlations with board gender diversity, more specific research about CG performance is needed. Concerning the quantity and duration of power outages, there is a positive correlation between male directors and executive members, implying that power interruption management must be targeted per function and directly linked to SAIDI and SAIFI metrics. As a result, the performance of executives can be monitored. The latter also entails the need for the executives to comply with managerial and engineering qualifications. This observation is also supported by a negative correlation between SAIDI and EBITDA, the independent board members, and the board’s gender diversity. Furthermore, no correlation concerning SAIFI also implies the need for relevantly qualified executives (and their composition) and the specification of their duties subject to SAIDI and SAIFI.

In this regard, the negative correlation of NPS with the size of the audit committee signals the need for a relevant level of formalisation to ensure customer satisfaction and, hence, contribute to their loyalty. A minor or no correlation between NPS and other variables tested also indicates the need for CG responsibilities specifically tailored to NPS.

As to the findings from regression analysis, it can be stated that seeking CG efficiency, the presence of executives and independent members, and the audit meeting frequency must be balanced. The positive correlation between the latter variables and the revenues, the net income, and EBITDA underscores the importance of the relevant competencies of the board members. EBDE’s negative correlation with the audit committee size also confirms the need for the auditors to provide the appropriate input. However, as to the negative correlation of NPS with the presence of executives, there is a need to align profitability and customer satisfaction goals. Nevertheless, the assessment of the governance structures and decision-making processes must be systematic due to the business dynamics, leading to the revaluation of the evolving challenges and possible solutions. Concerning variables regarding the number and duration of power outages, more specific and regular assessments are needed, with a focus on industry-specific challenges related to power system disruptions. Also, male and female directors and board gender diversity did not emerge as predictors in the models (similar to the Pearson correlation analysis results), suggesting further investigation.

Having in mind the above findings, the author notes that the existing audits for compliance with accounting standards as the standard practice are not sufficient for ensuring the effectiveness of CG. There is a need to improve CG accountability for decisions made by linking the performance of every decision-maker to concrete indicators per critical functions and operations. This would contribute to the transparency and compatibility of the results, allowing a more comprehensive assessment of the effectiveness of CG in terms of every respective authorised officer. Therefore, the principal shareholders of state-owned power suppliers should enact a mandatory procedure for conducting regular management audits. This will enhance the enterprises’ strategic, tactical, and operational management. As a result, CG could continuously focus on KPIs, paying due attention to SAIDI, SAIFI, and NPS (which, in principle, are found to have a fragmented correlation with CG).

Concerning privately owned energy producers, the research findings are subject to their applicability to state-owned entities. Nevertheless, these findings could serve as inputs for future comparative studies, such as those evaluating the sustainable and competitive performance of privately owned energy producers and state-owned entities. It is important to note that the research is limited to the period of 2019–2023. Therefore, the findings and insights are relevant to the business environment during that specific timeframe and may not be universally applicable to other periods. Furthermore, the findings are based on observational data, which constrains the ability to establish causal relationships. Therefore, caution should be taken when generalising correlations, as contextual factors and industry-specific dynamics could impact the observed correlations.

6. Conclusions

The intricate interactions of CG and performance results of the selected top five state-owned energy producers in Denmark, Estonia, Latvia, Lithuania, and Sweden were investigated.

By conducting a Pearson correlation analysis, significant correlations between CG variables and performance indicators were revealed. Notably, board size positively correlated with revenue and net income, highlighting the potential impact of board composition on financial results. Additionally, mixed correlations with board gender diversity were found, prompting further exploration concerning CG performance. An examination also signalled the importance of addressing the qualifications of male directors and executive members (and the board composition in this respect) regarding power interruption and managerial efforts to minimise the frequency and duration of power outages. Also, CG’s duties must be concrete when challenging the benchmarks of SAIDI, SAIFI, and NPS.

In analysing regression results, it is vital to consider the balance required between seeking CG efficiency, involving executives and independent members, and the frequency of audit meetings. The positive correlation between these variables and financial metrics underscores the importance of board members’ competencies. Also, the negative correlation of NPS with the presence of executives highlights the necessity to align profitability and customer satisfaction goals. It is crucial to systematically assess governance structures and decision-making processes to adapt to evolving business dynamics and challenges. Detailed, regular assessments are needed for industry-specific challenges related to power system disruptions. The regression results also showed the lack of significance of female directors and board gender diversity, warranting further investigation confirming the above conclusion of a Pearson correlation analysis.

Exploring the correlation between CG variables and SAIDI, SAIFI, NPS, and other KPIs (used across industries as the standard) of the investigated power suppliers, the results illuminated the multifaceted impact of governance structures on operational resilience and customer satisfaction. Noteworthy correlations emerged, suggesting that specific governance practices may influence financial metrics and an organisation’s operational and customer-centric aspects. For instance, the relevant presence of independent members is critical in avoiding conflicts of interest, according to a positive correlation with net income performance. Moreover, the findings signal the importance of a strong executive team for investigating power suppliers. Accordingly, this holds significance for other enterprises within the power sector to contemplate. In addition, the positive correlation between AMF, NEXE, and EBITDA observed must be treated as critical to strategic, tactical, and operational management objectives. In this regard, larger audit committees improve debt management and financial stability. And insignificant correlations of CG variables with SAIDI, SAIFI, and NPS also imply that CG performance assessment must be dynamic and linked to specific business functions and operations KPIs critical for sustainable and competitive performance across enterprises’ strategic, tactical, and operational management levels.

In the context mentioned above, it is imperative to acknowledge that CG variables exert a complex influence, intertwining with financial, operational, and customer-centric dimensions. The latter may serve as focal points for future research on CG and its impact on organisational performance regarding meta-flexibility, sustainable development, and competitiveness when transforming the energy sector and other sectors towards economies of renewable energy-based business models.