Determinants of Environmental, Social, and Governance Measures: Evidence from European Insurance Companies

Abstract

The present study aims to assess the determinants of the environmental, social, and governance (ESG) performance of European insurance companies, exploring the connections between the determinants and each of the three ESG pillars. Using a European panel dataset, the study includes a sample of the 30 most relevant European insurance companies listed in the stock exchange index Stoxx Europe 600 index, over the period 2016–2020. Main findings indicate that older insurance companies, with a larger dimension and profitability, less leveraged, with smaller boards, but with more gender diverse boards, with a sustainability committee and with a standalone sustainability report, tend to present a higher level of ESG performance. Findings also indicate that insurance companies with higher ESG performance are located in countries with higher gross domestic product growth rate. The current research setting has never been studied hitherto. Therefore, it adds to the limited empirical research on the determinants of ESG performance among insurance companies, which is focused on insurances companies in the United States of America.

1. Introduction

The integration of environmental, social, and governance (ESG) factors into corporate strategies has emerged as a critical benchmark for evaluating firms’ commitment to sustainable development (Friede et al., 2015; Buallay, 2019). Despite this growing emphasis, the academic literature remains limited in its analysis of ESG determinants within sector-specific contexts, particularly within the European insurance industry (Brogi et al., 2022). This study addresses this gap by investigating which firm-specific attributes influence ESG performance among European insurance companies.

The main objective of this research is to identify and examine the key determinants that influence ESG performance in the European insurance sector. Specifically, the study assesses how internal governance structures and financial characteristics relate to overall ESG scores and to each of its three dimensions—environmental, social, and governance. The central research problem lies in the lack of empirical evidence on ESG drivers in the European insurance industry, a sector that plays a pivotal role in the EU economy as its largest institutional investor (Brogi et al., 2022; Insurance Europe, 2020).

This research is guided by two primary questions: (1) What are the determinants of ESG performance in European insurance companies? and (2) which firm-level variables are associated with each ESG pillar—environmental, social, and governance? By addressing these questions, the study provides a more granular understanding of how ESG practices are operationalized in the insurance sector.

This research is motivated by two main aspects. First, inspired by the work of Brogi et al. (2022) on U.S. insurers, which called for similar studies across other geographic contexts, this article seeks to fill that empirical void in the European setting: a research setting never studied hitherto. Second, the European insurance sector is among the most developed globally, with a significant presence in both life and non-life insurance. According to the OECD (2024), in 2023, gross written premiums increased by an average of 12.4% in nominal terms and 6.2% in real terms—double the real growth rate observed in 2022 (OECD, 2024). The European insurance sector is also very resilient. The sector operates under the Solvency II regulatory framework, which imposes stringent capital and risk management requirements. Recently, the European Insurance and Occupational Pensions Authority (EIOPA) issued guidance aimed at harmonizing the supervision of captive insurers, promoting a proportionate and risk-based approach across the EU. In 2022, the average solvency ratio remained well above regulatory thresholds. Profitability also improved, with most insurers reporting positive investment returns, supported by a recovery in equity markets and stabilization in interest rates (IAIS, 2023). The EIOPA (EIOPA, 2023) documents that European insurers managed substantial investment portfolios, amounting to approximately EUR 8.57 trillion, in the first quarter of 2023. Asset allocation is primarily concentrated in sovereign and corporate bonds, equities, and investment funds, which collectively represent 88% of total holdings. This investment strategy seeks to strike a balance between stable returns and prudent risk management, especially in the context of market volatility (EIOPA, 2023). However, climate change poses an increasingly significant challenge to the European insurance sector. Extreme weather events—such as floods and wildfires—have resulted in substantial economic losses, of which only around 25% are currently insured (Financial Times, 2024). The EIOPA has called for coordinated action to address these risks, including the development of risk-sharing schemes and the strengthening of adaptation strategies (EIOPA, 2023). Since European countries are global leaders in advancing sustainability efforts (Buallay, 2019), the study of the determinants of ESG performance among European insurance companies is relevant to fill a gap in the literature and provide useful insights for investors, managers, and policymakers interested in promoting more sustainable goals, enhancing corporate governance and sustainability practices, and designing strategies to promote ESG performance among European insurance companies.

Using a panel dataset comprising 150 firm-year observations from 30 insurance companies listed in the Stoxx Europe 600 index between 2016 and 2020, the authors apply a fixed-effects panel data model. This methodological choice mitigates endogeneity and controls for unobservable heterogeneity (Brown et al., 2011), thereby enhancing the robustness of the findings.

The main findings reveal that ESG performance is positively associated with board gender diversity, the presence of a sustainability committee, standalone sustainability reporting, firm size, profitability, firm age, and country’s macroeconomic environment. In contrast, board size and leverage are negatively associated with ESG performance.

This study makes several important contributions. First, it responds to calls for more geographically diverse ESG research by extending the scope of previous studies (e.g., Brogi et al., 2022) to the European context. Second, it deepens the ESG literature with sector-specific evidence, which is particularly pertinent given the insurance industry’s exposure to environmental and financial risk (Scordis et al., 2014). Third, it demonstrates that governance characteristics and organizational maturity are critical in shaping sustainability outcomes, which has practical implications for investors, managers, and policymakers. For instance, the findings suggest that firms with more diverse and specialized governance bodies are better equipped to implement effective ESG strategies, thereby enhancing firm value and stakeholder trust (García-Meca et al., 2005). Fourth, for investors, the findings provide actionable insights into which company attributes are associated with ESG performance, enabling better alignment of investment decisions with sustainability goals. Fifth, for managers, the findings are relevant to enhance ESG strategies by adopting practices like establishing sustainability committees or producing standalone sustainability reports. Finally, for regulators, the study emphasizes the importance of supportive legal and economic environments, advocating for tailored policies that incentivize ESG performance.

The remainder of this article is organized as follows: Section 2 presents a review of the relevant literature and develops the research hypotheses. Section 3 details the research design, including the sample and econometric model. Section 4 presents the empirical results and discusses their implications. Section 5 concludes with key findings, contributions, and directions for future research.

2. Literature Review

While corporate social responsibility (CSR) represents a company’s moral commitment to society—regulating its internal activities in a way that aligns profit-making with generating positive economic, social, and environmental impacts (Sheehy & Farneti, 2021)—sustainability within organizations is defined as the adoption of strategies that not only meet the needs of the business and its stakeholders, creating value, but also safeguard the resources necessary for future generations, in accordance with ethical standards. More recently, the concept of ESG has emerged, aiming to quantify a range of non-financial information through key performance indicators that are relevant to investors and other stakeholders. This approach enhances the comparability of corporate social responsibility practices, supports more effective evaluation of non-financial performance, and thereby enables a clearer assessment of an organization’s broader impact on society and the environment (GPW, 2021).

A considerable body of the literature has examined the firm-level determinants of sustainability, CSR, and ESG reporting.

2.1. Determinants of Sustainability and CSR Reporting

The most consistent predictors of sustainability and CSR reporting across these studies include company size, profitability, leverage, and firm age.

Company size emerges as the most frequently cited determinant. Larger firms are more visible and face greater public scrutiny, which increases pressure to disclose non-financial information to legitimize their operations (Branco & Rodrigues, 2008; Hahn & Kühnen, 2013). This aligns with legitimacy theory and suggests that disclosure serves as a strategic tool for maintaining stakeholder trust.

Firm age is another determinant positively associated with sustainability and CSR reporting. Older firms are more established, have more developed reporting systems, and are typically more responsive to stakeholder demands (Owusu-Ansah, 1998; Al-Shammari et al., 2008).

Profitability is generally found to have a positive effect on sustainability reporting, as more profitable firms possess the financial resources to support ESG initiatives and bear the costs of disclosure (Nguyen & Nguyen, 2020; Rahman & Alsayegh, 2021). However, some studies report inconclusive or even negative relationships, which may be due to sectoral differences or variations in stakeholder pressure (Kumar et al., 2023; Bhatia & Tuli, 2017).

Leverage presents mixed results. Some research suggests that highly leveraged firms are more inclined to disclose information to reassure creditors and reduce agency costs (L. Oliveira et al., 2006; Hummel & Schlick, 2016), while others argue that firms with elevated financial risk may avoid voluntary disclosure to conceal their vulnerability (Ntim et al., 2013; Linsley & Shrives, 2006).

2.2. Determinants of ESG Reporting and Performance

More recent studies have focused on ESG reporting, an evolution of earlier CSR and sustainability frameworks. As with CSR, company size and profitability remain key predictors. Larger firms have greater resources and are more exposed to stakeholder expectations, making them more likely to adopt and report ESG practices (Buallay, 2019; Crespi & Migliavacca, 2020). Similarly, profitable firms can better absorb the costs of ESG initiatives and are more incentivized to disclose their performance to reinforce legitimacy (Brogi et al., 2022; Rahman & Alsayegh, 2021).

Other determinants include board composition and corporate governance mechanisms. Board diversity, particularly gender diversity, has been linked to enhanced ESG performance due to improved decision-making and oversight (Isidro & Sobral, 2015; Passas et al., 2022). The existence of a sustainability committee also contributes positively, as it signals institutional commitment and facilitates specialized oversight (Walls et al., 2012; Biswas et al., 2018). Managerial ability and the CEO’s education level positively influence voluntary ESG reporting (Kao et al., 2024), ESG performance (Oyindola, 2025), and ESG activities (Lewis et al., 2014; Liu et al., 2023).

The use of standalone sustainability reports is associated with higher ESG performance, as it reflects greater transparency and accountability (Zhang et al., 2021). However, caution must be exercised in interpreting ESG disclosures, as not all reports reflect genuine commitment—some may serve reputational or greenwashing purposes (Hummel & Schlick, 2016; Tsang et al., 2022).

Country-specific institutional and macroeconomic factors also influence ESG reporting. Firms operating in economically stable or legally stringent environments are more likely to engage in ESG disclosure (J. L. Campbell, 2007; Baldini et al., 2018).

2.3. ESG Determinants in Insurance Companies

Despite the insurance industry’s critical role in managing risk and investing in sustainability projects, studies focusing specifically on insurers are limited. This gap is particularly striking given that insurers are among the largest institutional investors in the European Union (Insurance Europe, 2020).

The few existing studies suggest that insurers engaged in CSR tend to produce higher-quality sustainability reports and are more proactive in ESG disclosure (Dropulić & Cular, 2019; Khovrak, 2020). Brogi et al. (2022) highlight that in U.S. insurance firms, profitability, firm size, and governance features such as board diversity significantly influence ESG performance. However, these findings cannot be directly generalized to the European context due to regulatory and market structure differences.

Given the limited sector-specific research in Europe, this study addresses a key gap by investigating the firm-level determinants of ESG performance in European insurance companies. In doing so, it extends prior research both geographically and contextually, offering new empirical evidence for scholars and practitioners concerned with sustainability in the insurance services sector.

3. Development of Hypotheses

3.1. Board Size

There is no consensus regarding the relationship between the number of board members and ESG reporting. However, larger boards are more likely to include members with expertise in accounting or finance, which is beneficial for CSR reporting. According to agency theory, larger boards provide better monitoring to reduce information asymmetry, which can positively influence the disclosure of voluntary information (Elzahar & Hussainey, 2012). This argument is supported empirically by studies such as Kathy Rao et al. (2012), Ntim et al. (2013), Rouf (2017), and Samaha et al. (2015), which found a positive relationship between board size and voluntary reporting.

Jensen (1993) presents an opposing view within agency theory, suggesting that larger boards are harder to monitor, potentially leading to a lack of coordination and communication. Smaller boards may improve a company’s performance by fostering better internal organization and more effective communication—two critical aspects of CSR disclosure (Kathy Rao et al., 2012). Improved internal coordination can enhance a company’s environmental performance, leading to increased disclosure (Ntim et al., 2013). Conversely, Bouaziz (2014) found a negative relationship between board size and voluntary reporting.

Given that most studies have identified a positive association between board size and ESG reporting, the following hypothesis is proposed:

Hypothesis 1 (H1). 

Board size is positively associated with ESG performance.

3.2. Gender Diversity in the Board of Directors

Women continue to be underrepresented in leadership positions. Passas et al. (2022) suggest that women play a relevant role in managing corporate crises, developing strategies, and solving problems. Women are crucial for combating fraud and fostering diversity, not as a replacement for men in leadership roles, but as part of a balanced and complementary integration of both genders in organizational leadership.

Isidro and Sobral (2015) argue that the primary way women on boards add value to a company is by enhancing the efficiency of the board and contributing unique skills and differentiated resources to the organization. According to these authors, gender diversity in the board of directors does not directly impact organizational performance but has an indirect positive effect by adding value to the company. This perspective aligns with agency theory, which posits that the role of directors is to manage the company efficiently, generating economic benefits for the organization and its investors. If, as suggested, the presence of women optimizes board performance, it can be inferred that both the company and its investors stand to benefit from female representation on the board (Gillan, 2006), basically because their presence in the board of directors promote accountability and transparency, two qualities integral to ESG reporting (Passas et al., 2022).

As a result, it is expected that gender diversity within a company is positively related to the implementation and reporting of ESG measures.

Hypothesis 2 (H2). 

The proportion of women on the board of directors is positively associated with ESG performance.

3.3. Board Independence

The board of directors typically comprises both internal and external directors. According to agency theory, managers rely on funds from investors, while investors depend on the human capital provided by managers to generate profits. As investors delegate some level of control to managers, there is a risk that managers may begin making decisions that serve their own interests, thereby increasing information asymmetry (Shleifer & Vishny, 1997).

Independent directors are more effective in enhancing investor returns because, unlike internal directors, they do not manage the company and have no personal incentives to act in self-interest. Their sole connection to the company lies in their governance role, which positions them as moderators in decision-making and conflicts arising from agency problems. They are also highly skilled in decision control (Fama & Jensen, 1983; Lim et al., 2007).

Independent directors play a crucial role in a company’s CSR initiatives by balancing the organization’s legitimacy with its social policies, thereby positively influencing voluntary reporting (Lim et al., 2007; Sharif & Rashid, 2014) and, consequently, the company’s performance.

Hypothesis 3 (H3). 

The proportion of independent directors is positively associated with ESG performance.

3.4. Sustainability Committee

According to resource dependence theory, the board of directors should be viewed as a provider of resources, offering advice and expertise, strengthening the company’s legitimacy and image, and fostering relationships with stakeholders. The board is also tasked with mitigating and addressing the organization’s environmental concerns (Hillman & Dalziel, 2003). By delegating sustainability matters to a specialized committee, directors are expected to gain better guidance in improving the organization’s environmental performance.

From the perspective of agency theory, having a CSR committee to handle corporate sustainability data more accurately can reduce information asymmetry and conflicts of interest between agents and principals. Additionally, it enhances the corporate governance of firms (Berrone & Gomez-Mejia, 2009) and a company’s value.

Conversely, Rodrigue et al. (2013) argue that sustainability committees are often restricted to ensuring compliance with sustainability standards, with their non-binding recommendations ultimately subject to the board’s approval. This limitation reduces their influence, making such committees appear as a tool for companies to showcase sustainability to stakeholders, even when they are not genuinely committed to it. Companies unwilling to invest in improving their ecological footprint might create sustainability teams as a facade to appear environmentally responsible (Berrone and Gomez-Mejia, 2009).

Empirical studies by Biswas et al. (2018), Mallin and Michelon (2011), Orazalin (2020), Spitzeck (2009), and Walls et al. (2012) have found that the presence of a sustainability committee is positively associated with a company’s social and environmental performance across various geographic regions. However, Berrone and Gomez-Mejia (2009) and Rodrigue et al. (2013) found no significant association between sustainability committees and environmental performance.

Hypothesis 4 (H4). 

The presence of a sustainability committee is positively associated with ESG performance.

3.5. Standalone Sustainability Report

In European countries, there is a greater focus on stakeholders compared to the United States, which is more oriented towards the needs of investors. Given this stakeholder-oriented approach, operational decisions in European companies are influenced by the moral values of consumers, employees, governments, and others (Dhaliwal et al., 2011). Maignan (2001) highlights that consumers in France and Germany place greater importance on companies’ ethical and social responsibilities, relegating economic responsibilities to a secondary role.

The disclosure of non-financial information by a company is, in principle, an indicator that it complies with sustainability measures. Since such reporting is not constrained by page limits and is directly related to CSR measures, companies can justify their practices more consistently and in greater detail. However, this also increases their exposure. Consequently, companies genuinely committed to societal and environmental concerns are more inclined to provide detailed accounts of their sustainability practices in standalone non-financial reports than companies engaging in greenwashing (Zhang et al., 2021).

The decision to produce such a report signals a company’s effort to be transparent. These reports are not only more detailed and easier to understand than financial reports but also provide information to assess the company’s long-term sustainability, with positive impacts on the company’s value. As a result, standalone sustainability reports are frequently examined by researchers in the field of sustainability (Dhaliwal et al., 2011; Wang & Li, 2016).

Hypothesis 5 (H5). 

Companies producing standalone sustainability reports, separate from management reports, are positively associated with ESG performance.

3.6. Size

Larger companies are under constant social pressure to make responsible decisions, as they are perceived to have sufficient resources to improve society. This requires ensuring that their activities go beyond serving the interests of investors to also benefit the environment and society, thereby seeking social legitimacy (Branco & Rodrigues, 2008).

Society supports a company’s operations if it demonstrates that its primary goal is not solely maximizing profit but also fulfilling the needs and expectations of its stakeholders. Larger companies are, therefore, more likely to disclose non-financial matters, as they not only have the resources to allocate but are also subject to greater scrutiny from their stakeholders. This aligns with legitimacy theory (Deegan, 2002). In contrast, smaller companies often lack the resources required for such disclosures, which are associated with higher costs. They also have fewer stakeholders and a smaller environmental or social impact (Hahn & Kühnen, 2013).

According to the agency theory, larger companies face greater information asymmetry. By disclosing more information, these companies can reduce this asymmetry, minimize conflicts of interest, and reassure investors, ultimately lowering agency costs (García-Meca et al., 2005) and improving the company’s value.

Several studies, including Branco and Rodrigues (2008), Brogi et al. (2022), Hahn and Kühnen (2013), and Rahman and Alsayegh (2021), have demonstrated a positive relationship between company size and voluntary disclosure.

Hypothesis 6 (H6). 

Company size is positively associated with ESG performance.

3.7. Profitability

ESG measures and their associated reporting improve organizational profitability (Buallay, 2019; Friede et al., 2015; Orlitzky et al., 2003). Highly profitable companies may face greater public scrutiny, consistent with legitimacy theory, or even political exposure (Bewley & Li, 2000; Crespi & Migliavacca, 2020). As a result, these companies are incentivized to explain how their profitability is achieved to legitimize themselves in the eyes of stakeholders, as profit can often be associated with worker exploitation and environmental degradation. Consequently, companies with higher profitability are also encouraged to disclose whether they operate in accordance with environmental and societal norms, avoiding profit at any cost, or whether they allocate part of their profit to community-related initiatives (Gamerschlag et al., 2011).

However, the relationship between profitability and disclosure is not universally agreed upon. Several authors have found no significant association between profitability and the disclosure of information (Chih et al., 2010; Hahn & Kühnen, 2013).

On the other hand, in line with the theoretical framework, empirical studies such as Brogi et al. (2022) and Rahman and Alsayegh (2021) have identified a positive relationship between profitability and sustainability reporting.

Hypothesis 7 (H7). 

Profitability is positively associated with ESG performance.

3.8. Leverage

Companies with high levels of leverage tend to produce higher-quality sustainability reports. This is expected, as the lenders of these companies require more detailed information and control as leverage increases, to assess the company’s ability to repay its debts. Consequently, these companies are encouraged by their debtholders to disclose more information beyond financial reports, thereby reducing information asymmetry and demonstrating legitimacy to stakeholders (Branco & Rodrigues, 2008; Hummel & Schlick, 2016; Linsley & Shrives, 2006). Similarly, from the perspective of agency theory, highly leveraged companies have greater incentives to voluntarily disclose information to reduce agency costs (L. Oliveira et al., 2006). These arguments are consistent with findings of prior studies (Hummel & Schlick, 2016; Rahman & Alsayegh, 2021).

Conversely, Linsley and Shrives (2006) argue that organizations with higher risk, such as those with elevated leverage levels, may resist emphasizing their risk exposure and may be reluctant to disclose information voluntarily. No significant association was found between leverage and voluntary disclosure in Depoers (2000). Furthermore, Ntim et al. (2013) identified a negative association between leverage levels and non-financial reporting.

Hypothesis 8 (H8). 

Leverage is associated with ESG performance.

3.9. Age

According to Owusu-Ansah (1998), newer companies may face competitive disadvantages if they disclose excessive information, as they are still in the process of studying the market and developing their products. Conversely, older companies do not face this issue and, having been in operation for a longer period of time, they are incentivized to demonstrate conscientiousness in their strategies. This leads to a greater motivation to disclose information (Owusu-Ansah, 1998; Al-Shammari et al., 2008).

Hypothesis 9 (H9). 

Age is associated with ESG reporting.

3.10. Country

The way stakeholders evaluate a company depends on various factors, including ethical considerations influenced by culture, legal frameworks, and economic conditions. According to institutional theory, countries represent distinct social environments composed of normative, cognitive, and regulative elements, which collectively contribute to the legitimization of companies by stakeholders (Scott, 1995).

J. L. Campbell (2007) argues that companies being less likely to adhere to social norms when achieving substantial short-term profits is unlikely due to economic instability within their country. Such limitations may arise from factors like low consumer purchasing power or high levels of inflation (J. L. Campbell, 2007; Orlitzky et al., 2003). J. L. Campbell (2007) also suggests that companies in countries with legal systems enforcing responsible corporate behavior are more likely to act conscientiously, particularly when these norms result from a consensus among governments, stakeholders, and businesses.

Hypothesis 10 (H10). 

Country is associated with ESG performance.

4. Research Design

4.1. Sample

From the 600 companies included in the Stoxx 600 Europe Index, our initial sample comprises 31 insurance companies selected as of 31 October 2022. We chose the STOXX Europe 600 because it is a broad, market-capitalization-weighted stock index that includes 600 publicly traded companies across 17 European countries, covering large, mid, and small capitalizations (STOXX, 2022). It offers substantial sectoral and geographical diversification and represents approximately 90% of the free-float market capitalization of the European stock market (STOXX, 2022). The index includes firms from major economies such as the United Kingdom, Germany, France, and Switzerland, thereby capturing a wide range of institutional settings, regulatory frameworks, and corporate governance systems (Brogi et al., 2022). Due to its breadth and representativeness, the STOXX Europe 600 is frequently used in accounting and ESG research. It provides a robust benchmark for studying cross-country variation in sustainability practices, financial transparency, and non-financial reporting (Buallay, 2019; Crespi & Migliavacca, 2020). Moreover, the inclusion of listed firms ensures the availability and comparability of ESG and financial data, which are essential for conducting empirical research with strong external validity.

Table 1. Sample selection.

Panel A: Sample selection criteria		Number of companies	Firm-year observations
	Companies included in the Stoxx 600 Europe index	600	3000
	Investment funds	−3	−15
	Companies that do not belong to insurance industry	−566	−2830
	Companies without data in the Eikon Databse	−1	−5
	Final sample	30	150
Panel B: Sample distribution by insurance branch		Number of companies	Percentage
	Life.insurance	10	33%
	Non life-insurance	20	67%
	Final sample	30	100%
Panel C: Sample distribution per country		Number of companies	Percentage
	Germany	3	10%
	Belgium	1	3%
	Denmark	2	7%
	Finland	1	3%
	France	2	7%
	Netherlands	3	10%
	Italy	2	7%
	Norway	2	7%
	Poland	1	3%
	United Kingdom	8	27%
	Switzerland	5	17%
	Final sample	30	100%

details the selection criteria.

First, to isolate the most relevant European insurance companies, we excluded 3 investment funds and 566 non-insurance companies, leaving 31 insurance companies. One insurance company was further excluded from the initial sample due to a lack of required data in the Eikon database. The final sample comprises 30 insurance companies from 11 European countries for the period 2016–2020, yielding a total of 150 observations. The study began in 2016, as this is the first year for which all necessary data were available for the 30 companies.

The insurance sector is the largest investor in the EU economy, with investments amounting to EUR 10.627 trillion by 2020, equivalent to 61% of the European GDP. As a critical industry for economic growth, insurers are in a prime position to invest in sustainability-related projects, such as resource efficiency and carbon neutrality. Many insurers have voluntarily disclosed sustainability-related matters, developed sustainability standards, and invested in related projects, all aligned with ESG strategies (Insurance Europe, 2020).

Table 1 (Panel B) categorizes the sample into Life and Non-Life branches, showing that one-third of the insurance companies in the final sample belong to the Non-Life branch, while the remaining two-thirds are from the Life branch.

When distributed by country, most companies are from the United Kingdom (8), Switzerland (5), the Netherlands (3), and Germany (3), with the remaining 11 companies spread across seven other European countries.

4.2. Econometric Model

To analyze the determinants of ESG performance in European insurance companies, including fixed effects, the following econometric model was defined (Equation (1)):

$$\mathbf{E}\mathbf{S}\mathbf{G}\mathbf{i}\mathbf{t}=\mathsf{\beta }0+{\sum }_{\mathit{i}=1}^{\mathit{n}}\mathsf{\beta }\mathit{i}\mathit{t}\mathit{D}\mathit{e}\mathit{t}\mathit{e}\mathit{r}\mathit{m}\mathit{i}\mathit{n}\mathit{a}\mathit{n}\mathit{t}\mathit{s}\mathit{i}\mathit{t}+\mathit{\gamma }\mathit{i}+\mathsf{\epsilon }\mathit{i}\mathit{t}$$

(1)

To assess the dependent variable “ESG performance”, this study employs ESG scores provided by the Eikon database (formerly Thomson Reuters, now part of LSEG), a widely used and reputable data source in academic and professional ESG research. Eikon ESG scores are particularly valuable due to their comprehensive coverage, consistent methodology, and transparency. The database covers over 10,000 companies globally and uses more than 450 company-reported data points to generate ESG scores, which are categorized into environmental, social, and governance pillars and aggregated into an overall ESG score. These metrics are constructed based on publicly disclosed information from annual reports, sustainability reports, and third-party sources, thus reducing subjectivity and enhancing comparability across firms and over time (Chatterji et al., 2016; Liang & Renneboog, 2017).

Moreover, Eikon applies a materiality-weighted scoring model, which accounts for industry-specific relevance of ESG issues, aligning with the Sustainability Accounting Standards Board (SASB) framework and the Global Reporting Initiative (GRI) principles. Its transparent scoring methodology has led to its growing adoption in empirical studies on corporate sustainability, non-financial disclosure, and ESG-related investment decisions (Brogi et al., 2022; Buallay, 2019). The choice of Eikon, thus, ensures that ESG performance is measured using a systematic and widely recognized framework, enhancing the reliability, validity, and replicability of the results.

Companies differ in terms of challenges and choices over time. This implies that ESG performance, activities, policies, or even reporting may be jointly and dynamically determined by unobserved company-specific heterogeneities and time-related variations, such as managerial talent, corporate culture, and complexity (Henry, 2008), which simple OLS regression may be unable to detect (Gujarati, 2003; Wooldridge, 2010). Given the panel nature of our data, we conduct our analyses by using panel data regression techniques to mitigate potential endogeneities associated with sample selection bias and omitted variables bias related to unobserved heterogeneity of a company-specific and/or time-invariant nature (Brown et al., 2011).

Preliminary diagnostic tests corroborate the use of panel data regression techniques through a fixed-effects model: the F statistic (=6.533; p-value < 0.01) rejects the use of the pooled model; the Hausman test (=25.280; p-value < 0.01) rejects the use of the random-effects model; and the Breusch–Pagan test (=55.628; p-value < 0.01) corroborates the use of a fixed-effects model. Additionally, preliminary tests indicate that there is not enough evidence to conclude that there is a structural break in the regression model (Chow test = 0.772; p-value = 0.656), there is no significant evidence of cross-sectional dependence (Pesaran’s test = 0.460; p-value = 0.6457), and the series in panel data are stationary (Levin–Lin–Chu test = −6.234; p-value = 0.000).

The independent variables, measurement criteria, and predicted signs are described in

Table 2. Definition of variables.

Varibles	Measurement	Predicted Signal	References
Board size	Number of directors in the board of directors	+	Elzahar and Hussainey (2012); Kathy Rao et al. (2012); Ntim et al. (2013); Rouf (2017); Samaha et al. (2015).
Gender diversity	Proportion of women in the board of directors	+	Adams and Ferreira (2009); K. Campbell and Mínguez-Vera (2008); Hafsi and Turgut (2013); Isidro and Sobral (2015).
Independent directors	Proportion of independent directors in the board of directors	+	Ntim et al. (2013); Rouf and Hossan (2021); Sharif and Rashid (2014).
Sustainability committee	Dummy = 1 if the insurance company has a sustainability committee, =0 otherwise.	+	Biswas et al. (2018); Mallin and Michelon (2011); Orazalin (2020); Spitzeck (2009); Walls et al. (2012).
Standalone sustainability report	Dummy = 1 if the insurance company issues a standalone sustainability report, =0 otherwise.	?	Dhaliwal et al. (2012)
Size	Natural logarithm of total assets	+	Branco and Rodrigues (2008); Brogi et al. (2022); Elzahar and Hussainey (2012); Hahn and Kühnen (2013); Rahman and Alsayegh (2021).
Profitability	Return On Assets: earnings before taxes to total assets	+	Brogi et al. (2022); Chih et al. (2010); Hahn and Kühnen (2013); Rahman and Alsayegh (2021).
Leverage	Total debt to total assets	?	Depoers (2000); Hummel and Schlick (2016); Ntim et al. (2013); Rahman and Alsayegh (2021).
Age	Number of years old of the companies since inception	?	J. Oliveira et al. (2011).
Country	GDP growth rate	?	J. Oliveira et al. (2019).

A positive sign (+) indicates that this variable is expected to have a positive effect. The ? means that there is no expectation regarding the signal of the effect.

. Consistent with the prior literature, the variable “board size” as measured by the number of directors in the board (Elzahar & Hussainey, 2012; Kathy Rao et al., 2012; Ntim et al., 2013; Rouf, 2017; Samaha et al., 2015). The variable “gender diversity” was proxied by the proportion of women in the board of directors (Adams & Ferreira, 2009; K. Campbell & Mínguez-Vera, 2008; Hafsi & Turgut, 2013; Isidro & Sobral, 2015). The variable “independent directors” was assessed by the proportion of independent directors in the board (Ntim et al., 2013; Rouf & Hossan, 2021; Sharif & Rashid, 2014).

The variable “sustainability committee” was measured by a dummy variable that was assigned 1 if the company has a sustainability committee and 0 otherwise (Biswas et al., 2018; Mallin & Michelon, 2011; Orazalin, 2020; Spitzeck, 2009; Walls et al., 2012). The variable “standalone sustainability report” was assessed by a dummy variable that was assigned 1 if the company issues a standalone sustainability report and 0 otherwise (Dhaliwal et al., 2012). The variable “size” was proxied by the natural logarithm of total assets (Branco & Rodrigues, 2008; Brogi et al., 2022; Elzahar & Hussainey, 2012; Hahn & Kühnen, 2013; Rahman & Alsayegh, 2021). The variable “profitability” was measured by the return on assets ratio computed by earnings before taxes to total assets (Brogi et al., 2022; Chih et al., 2010; Hahn & Kühnen, 2013; Rahman & Alsayegh, 2021). The variable “leverage” was assessed by total debt to total assets (Depoers, 2000; Hummel & Schlick, 2016; Ntim et al., 2013; Rahman & Alsayegh, 2021). The variable “age” was proxied by the number of years old of the companies since their inception (J. Oliveira et al., 2011). The variable “country” was assessed by the gross domestic product growth rate of each country in a specific year (J. Oliveira et al., 2019).

Data regarding the dependent and independent variables were extracted from the Eikon database. The data on gross domestic product growth rate was extracted from the World Bank Open Data.

5. Results and Discussion

5.1. Descriptive Statistics

Table 3. Descriptive statistics.

Variable	N	Minimum	Maximum	Mean	Standard Deviation
ESG Combined score	150	25.670	91.470	61.763	15.030
Environmental pillar	150	15.090	98.900	63.799	24.589
Governance pillar	150	16.820	96.240	68.284	18.630
Social pillar	150	25.400	93.520	62.574	17.331
Size of the board of directors	150	4.000	21.000	11.227	3.041
Gender diversity	150	0.000	0.667	0.339	0.107
Independent directors	150	0.200	1.000	0.688	0.211
Size	150	15.131	20.758	18.614	1.468
Profitability	150	−0.023	0.108	0.020	0.023
Leverage	150	0.011	0.297	0.063	0.055
Age	150	8.000	217.000	98.933	66.387
Country	150	−11.031	5.945	0.429	3.606
		N	%
Sustainability committee	Dummy = 1	121	80.7
	= 0	29	19.3
Standalone sustainability report	Dummy = 1	144	96.0
	= 0	6	4.0

presents the descriptive statistics for each dependent and independent variable.

The dependent variable, ESG combined score, has an average of 61.763, indicating that more than half of the insurers have a good ESG performance. This score ranges from a minimum of 25.67 to a maximum of 91.47. However, the tendency does not deviate significantly from the mean, as evidenced by a standard deviation of 15.03. Among the ESG pillars, the governance pillar has the highest average of 68.284, followed by the environmental pillar with an average of 63.799, and the social pillar with an average of 62.574. The social pillar’s average is the closest to the ESG combined score, including its minimum and maximum values. Thus, while the governance pillar has a minimum of 26.82, it remains the most assured ESG component among European insurance companies on average, with a standard deviation of 18.63.

The size of the board of directors ranges from a minimum of 4 directors to a maximum of 21, with an average of 11 directors. The proportion of women on the board averages 0.339, varying between boards with no female representation and those where 0.667 of the members are women. On average, 0.688 of the board members are independent directors. The company size has an average of 18.614, with a minimum of 15.131 and a maximum of 20.758. The profitability ratio (ROA) has an average of 0.02, with some insurers reporting negative ROA values as low as −0.023 (minimum) and a maximum profitability of 0.108. A plausible explanation for the lower profitability levels is the inclusion of 2020 data in the study, during which the EU insurance sector incurred estimated losses of EUR 36 to EUR 72 billion due to the COVID-19 pandemic (Insurance Europe, 2020). The average leverage level of the companies is 0.063, with a minimum of 0.011 and a maximum of 0.297. The variable age indicates that European insurers have an average of 99 years of existence, with the newest company being 8 years old and the oldest 217 years old. The variable country shows an average of 0.0429, with a minimum of −0.11031 and a maximum of 0.05945. All negative GDP growth values, including the lowest value, present the direct effect of COVID-19. European GDP fell in 2020 for the first time since 2013, declining by approximately 5.9% (Insurance Europe, 2020). A total of 80.7% of the insurers in the study have a sustainability committee, and the vast majority (96%) produce a standalone sustainability report separate from the management report, with only six insurers not doing so.

5.2. Correlation Matrix

Table 4. Correlation matrix.

Variables		(1)		(2)		(3)		(4)	(5)		(6)		(7)		(8)		(9)	(10)		(11)
	Panel A: Pearson correlation for continuous variables
(1)	ESG combined score	1.000
(2)	Board size	0.189	**	1.000
(3)	Gender diversity	0.038		0.101		1.000
(4)	Independent directors	0.148		−0.178	**	−0.179	**	1.000
(5)	Size	0.588	***	0.200	**	0.017		−0.012	1.000
(6)	Profitability	−0.316	***	−0.189	**	0.247	***	−0.093	−0.606	***	1.000
(7)	Leverage	−0.108		0.033		0.187	**	−0.140	−0.173	**	0.213	***	1.000
(8)	Age	0.360	***	0.383	***	−0.160		0.051	0.423	***	−0.367	***	−0.026		1.000
(9)	Country	−0.027		−0.069		−0.140		0.017	0.015		0.109		−0.126		−0.029		1.000
	Panel B: Spearman correlation for categorical variables
(10)	Sustainability committee	0.295	***	0.085		0.002		−0.078	0.227	***	−0.085		−0.180	**	0.268	***	−0.084	1.000
(11)	Standalone sustainability report	0.236	***	−0.209	**	0.077		0.067	0.136		−0.032		−0.079		0.185	**	−0.021	0.417	***	1.000

Correlation significant at a level of: *** 0.01 (2-tailed), ** 0.05 (2-tailed).

presents the correlations between the independent variables and the dependent variable.

The dependent variable is positively correlated with board size (p-value < 0.05), company size, age, the presence of a sustainability committee, and standalone sustainability reporting (p-value < 0.01). It is also negatively correlated with profitability (p-value < 0.01). The highest variance inflation factor (VIF) value is 2.028, indicating low levels of multicollinearity (VIF < 10).

5.3. Regression Tests

The assumptions underlying the regression models were tested for autocorrelation, multicollinearity, heteroscedasticity, outliers and influential observations, and the normality of residuals. The continuous dependent and independent variables follow a normal distribution.

Table 5. Regression tests.

Variables	Predicted Signal	ESG Combined Score
		Fixed-Effects Model
Intercept		−409.012	***
		−(4.331)
Board size	+	−1.460	††
		−(2.325)
Gender diversity	+	14.569	†
		(1.315)
Independent directors	+	5.535
		(0.522)
Sustainability committee	+	4.336	†††
		(2.858)
Standalone sustainability report	?	6.499	***
		(2.709)
Size	+	14.535	†††
		(4.117)
Profitability	+	104.609	†
		(1.382)
Leverage	?	−38.654	**
		−(2.105)
Age	?	2.003	***
		(2.760)
Country	?	0.690	*
		(1.850)
Year fixed effects		Yes
Firm fixed effects		Yes
Model fit:
Adjusted R2		0.289
F statistics		58.971	***
Durbin–Watson		1.450
VIF		<2.028
Observations		150

Significant at a level: ††† 0.01 (1-tailed), †† 0.05 (1-tailed), and † 0.1 (1-tailed). Significant at a level: *** 0.01 (2-tailed), ** 0.05 (2-tailed), and * 0.1 (2-tailed). Standard errors are heteroscedasticity-adjusted and clustered at the firm level. The T statistics are in parentheses. A positive sign (+) indicates that this variable is expected to have a positive effect. The ? means that there is no expectation regarding the signal of the effect.

presents the results of panel data regression tests using a fixed effects approach.

The results shown in Table 5 demonstrate that the regression model is generally valid. The model is statistically significant (F = 58.971; p-value < 0.01). The explanatory power of the independent variables in relation to the variation in the ESG combined score is 28.9% (adjusted R² = 0.289).

Findings indicate that ESG performance is positively associated with gender diversity on the board (p-value < 0.1), the presence of a sustainability committee (p-value < 0.01), standalone sustainability reporting (p-value < 0.01), company size (p-value < 0.01), profitability (p-value < 0.1), company age (p-value < 0.01), and country (p-value < 0.1). These results support hypotheses H2, H4, H5, H6, H7, H9, and H10.

Regarding gender diversity, the findings align with the theory developed by Isidro and Sobral (2015), who argue that women on boards add value to companies, thereby improving their performance across all dimensions, including non-financial performance.

The positive association found between the presence of a sustainability committee and ESG performance is consistent with previous findings by Biswas et al. (2018), Mallin and Michelon (2011), Orazalin (2020), Spitzeck (2009), and Walls et al. (2012). These studies suggest that companies with sustainability committees address sustainability-related issues more precisely, leveraging greater expertise and specificity (Berrone & Gomez-Mejia, 2009). However, the result contrasts with the empirical findings of Berrone and Gomez-Mejia (2009) and Rodrigue et al. (2013), who found no significant relationship between the presence of a sustainability committee and ESG performance.

According to Zhang et al. (2021), companies that produce standalone sustainability reports, separate from their management reports, are generally more concerned with the environment and society, disclosing their sustainability practices in greater detail. This explains why standalone sustainability reporting is positively associated with the dependent variable.

The positive relationship between company size and ESG performance aligns with legitimacy theory, as larger companies are subject to greater public scrutiny and are, thus, pressured to disclose more information (Deegan, 2002). This finding also supports agency theory, as increased disclosure reduces information asymmetry (García-Meca et al., 2005).

The results regarding profitability are consistent with the empirical studies of Brogi et al. (2022) and Rahman and Alsayegh (2021), which found a positive relationship between profitability and non-financial reporting. This outcome can be explained by legitimacy theory, as more profitable companies have greater incentives to demonstrate that their profits are earned legitimately and that, in part, these profits are used to contribute to society and the environment (Bewley & Li, 2000; Crespi & Migliavacca, 2020; Gamerschlag et al., 2011).

As noted by Owusu-Ansah (1998), who concluded that older companies tend to disclose more information, the results in Table 5 indicate that age is a determinant of ESG performance.

The variable country, measured by GDP, showed a positive association with ESG performance. According to J. L. Campbell (2007), this can be explained by the fact that companies in more economically stable countries are likely to have greater motivation to disclose information, as the likelihood of achieving short-term profits makes them more willing to comply with social norms. However, this result contrasts with the findings of J. Oliveira et al. (2019), who did not identify a relationship between the two variables.

No relationship was found between board independence and ESG performance, failing to support Hypothesis H3.

Board size demonstrates a negative relationship (p-value < 0.05) with the dependent variable, failing to support Hypothesis H1, which had predicted a positive association between the two variables. This result is consistent with Bouaziz (2014) but contradicts the findings of Kathy Rao et al. (2012), Ntim et al. (2013), Rouf (2017), and Samaha et al. (2015). The negative association between these variables may be attributed to a lack of communication and coordination within larger boards (Jensen, 1993).

Regarding leverage, a negative association (p-value < 0.05) was found with the ESG combined score, supporting Hypothesis H8, which predicted a relationship between the two variables. This result is consistent with Ntim et al. (2013), who also identified a negative relationship between leverage and non-financial reporting. Companies with high leverage levels may resist voluntarily disclosing information to avoid drawing attention to their risk exposure (Linsley & Shrives, 2006). However, this finding is inconsistent with the studies of Hummel and Schlick (2016) and Rahman and Alsayegh (2021), which identified a positive association between leverage and ESG performance.

5.4. Additional Analysis

Table 6. Additional regression tests.

Variables	Predicted Signal	Environmental Pillar		Governance Pillar		Social Pillar
		Fixed-Effects Model		Fixed-Effects Model		Fixed-Effects Model
Intercept		−137.427		−317.762	***	−240.903	**
		−(1.025)		−(3.902)		−(2.120)
Board size	+	−0.396		−2.257	††	−0.498
		−(0.347)		−(2.306)		−(0.657)
Gender diversity	+	−11.254		41.668	†††	−3.529
		−(0.499)		(2.952)		−(0.333)
Independent directors	+	−1.918		5.896		11.894	††
		−(0.137)		(0.664)		(1.779)
Sustainability committee	+	−2.970		−3.400		8.792	†††
		−(0.500)		−(0.975)		(3.872)
Standalone sustainability report	?	19.984	*	10.535	**	6.640	**
		(1.404)		(2.068)		(1.707)
Size	+	24.609	†††	0.690		7.724
		(4.108)		(0.206)		(1.262)
Profitability	+	320.669	††	36.311		52.031
		(2.291)		(0.274)		(0.784)
Leverage	?	−89.714	*	−38.090	**	−19.745
		−(1.552)		−(1.812)		−(1.175)
Age	?	−2.675	**	3.787	***	1.465	***
		−(2.013)		(4.952)		(3.528)
Country	?	−0.384		0.039		0.327	**
		−(1.247)		(0.181)		(2.247)
Year fixed effects		Yes		Yes		Yes
Firm fixed effects		Yes		Yes		Yes
Model fit:
Adjusted R2		0.209		0.504		0.372
F statistic		6.641	***	7.868	***	4.370	***
Durbin–Watson		1.306		1.309		1.409
VIF		<2.028		<2.028		<2.028
Observations		150		150		150

Significant at a level: ††† 0.01 (1-tailed), and †† 0.05 (1-tailed). Significant at a level: *** 0.01 (2-tailed), ** 0.05 (2-tailed), and * 0.1 (2-tailed). Standard errors are heteroscedasticity-adjusted and clustered at the firm level. The T statistics are in parentheses. A positive sign (+) indicates that this variable is expected to have a positive effect. The ? means that there is no expectation regarding the signal of the effect.

presents the panel data regression models, using a fixed-effects approach, for the three ESG pillars: environmental, social, and governance.

Findings indicate that models with fixed effects are generally valid. These models are statistically significant (F-statistic for environmental = 7.635; F-statistic for social = 14.768; F-statistic for governance = 17.253; p-value < 0.01). The explanatory power of the independent variables in relation to the variation in environmental, social, and governance is 20.9% (adjusted R² = 0.209), 37.2% (adjusted R² = 0.372), and 50.4% (adjusted R² = 0.504), respectively.

These regression models reveal the following: board size is negatively associated with the governance pillar (p-value < 0.05) but shows no association with the environmental or social pillars. Gender diversity is positively associated with the governance pillar (p-value < 0.01) but shows no relationship with the environmental or social pillars. Although board independence was not found to be associated with the ESG combined score, it has a positive relationship with the social pillar (p-value < 0.05). The presence of a sustainability committee has a positive impact solely on the social pillar (p-value < 0.01). Standalone sustainability reporting shows a positive association with all three components: environmental (p-value < 0.1), social, and governance (p-value < 0.05). Company size only has a positive impact on the environmental pillar (p-value < 0.01) and shows no relationship with the other components. The same applies to profitability (p-value < 0.05). Leverage shows a negative relationship with the environmental and governance pillars (p-value < 0.1 and p-value < 0.05, respectively), but no relationship with the social pillar. Regarding age, there is an association with all three pillars: positive with governance and social (p-value < 0.01) and negative with environmental (p-value < 0.05). The country variable shows a positive relationship only with the social pillar (p-value < 0.05).

6. Conclusions

This study contributes to the expanding literature on ESG performance by identifying the firm-level determinants of ESG outcomes within the European insurance industry, a sector that, despite its strategic financial role, remains underexplored in ESG research. Drawing on a sample of 150 firm-year observations from 30 insurers listed in the STOXX Europe 600 index over the period 2016–2020, and using a fixed-effects panel data model, the study offers empirical insights into how organizational characteristics, governance structures, and contextual factors influence ESG performance.

The results indicate that ESG performance is positively associated with board gender diversity, the presence of a sustainability committee, the issuance of standalone sustainability reports, firm size, profitability, firm age, and the macroeconomic environment (as proxied by national GDP growth). Conversely, board size and leverage are negatively associated with ESG performance. These findings are broadly consistent with the prior literature that emphasizes the role of internal governance mechanisms and organizational maturity in fostering sustainable practices (Buallay, 2019; Brogi et al., 2022; Isidro & Sobral, 2015; Zhang et al., 2021).

Importantly, the study finds that board gender diversity significantly enhances ESG performance, echoing findings by Passas et al. (2022) and Isidro and Sobral (2015), who highlight the value women bring to board deliberations and strategic decision-making. The positive effect of sustainability committees and standalone reports reinforces earlier work (e.g., Walls et al., 2012; Biswas et al., 2018), suggesting that formal governance mechanisms and transparent reporting practices contribute to a firm’s ESG credibility.

The negative association between leverage and ESG performance supports agency-theory-based arguments that highly leveraged firms may avoid extensive non-financial disclosure to limit visibility of financial vulnerability (Linsley & Shrives, 2006; Ntim et al., 2013). The inverse relationship between board size and ESG performance lends credence to concerns about coordination inefficiencies in larger boards (Jensen, 1993), contrasting with the assumption that size inherently improves monitoring capacity (Kathy Rao et al., 2012).

This research makes several key contributions. First, it extends the geographical scope of ESG studies by focusing on European insurers, addressing the call by Brogi et al. (2022) for research in contexts beyond the United States. Second, it contributes sector-specific insights, recognizing the unique regulatory, investment, and risk management roles played by insurers in advancing sustainability. Third, the study empirically validates the relevance of both internal (governance-related) and external (macroeconomic) variables in shaping ESG outcomes. Fourth, it offers actionable implications: managers are encouraged to institutionalize ESG practices via board diversity and sustainability committees; investors are provided with cues to evaluate ESG performance; and policymakers gain evidence to support regulatory reforms aligned with sustainable finance objectives.

In terms of limitations, the analysis is restricted to listed insurers in the STOXX Europe 600 index, potentially excluding smaller or privately held insurers whose ESG behaviors may differ. Furthermore, the five-year period (2016–2020) may not fully capture long-term ESG strategy evolution, especially considering the disruptions caused by the COVID-19 pandemic. The study also relies on ESG scores from a single database (Eikon), which, while robust and widely used, may reflect certain methodological assumptions not shared across other ESG data providers (Chatterji et al., 2016).

Future research could expand this work in several directions. Comparative studies across financial sub-sectors (e.g., banks vs. insurers) or regions (e.g., Europe vs. Asia) would enhance understanding of contextual influences. Longitudinal studies over extended periods could capture temporal shifts in ESG integration, particularly in response to regulatory reforms such as the EU Taxonomy or the Corporate Sustainability Reporting Directive (CSRD). Additionally, qualitative approaches—such as case studies or interviews with ESG officers—could complement quantitative findings by revealing internal motivations, cultural factors, or institutional constraints that shape ESG performance beyond what numerical indicators capture.