This study aims to investigate the impact of governance structures on eco-innovation within insurance companies. It specifically examines the impact of various board-level and organizational characteristics, including board gender diversity, board meeting attendance, board size, board-specific skills, chief executive officer−chairman separation, corporate social responsibility (CSR) sustainability committee, environmental management team, independent board members and sustainability compensation incentives.
This research uses a sample of 252 insurance companies operating between 2011 and 2022, extracting from the Thomson Reuters Refinitiv Eikon database. It uses panel and pooled ordinary least squares regressions to analyze the relationship between governance variables and an eco-innovation index, which is developed using five proxies for eco-innovation practices.
The findings reveal that gender diversity on boards and board meetings enhances eco-innovation performance. In addition, the presence of dedicated sustainability structures − such as CSR committees and environmental management teams − and sustainability-linked compensation incentives positively correlates with eco-innovation scores. Independent board members also support eco-innovation, while financially oriented board skills show a negative association.
This study provides novel insights into the governance determinants of eco-innovation in the insurance sector, which remains underexplored in the existing literature. Building on the eco-innovation index developed by Albitar et al. (2023), and to the best of the authors’ knowledge, this paper is the first to apply the index within the insurance industry, using a global sample of firms. In doing so, it examines how corporate governance structures are linked to eco-innovation performance. Furthermore, the authors use survival analysis to assess the long-term impact of specific sustainability governance mechanisms − namely, the adoption of sustainability incentives, the presence of an environmental management team and CSR committees − on the likelihood of firms achieving and maintaining higher levels of eco-innovation.
1. Introduction
The increasing frequency and seriousness of climate-related negative events have underscored the need for improvement among eco-innovation. Firms have to manage new risks to face the rising of global temperatures and the extreme weather events. This challenge is widely recognized by regulators, policymakers and industry professionals: in fact, new regulatory frameworks aimed at integrating sustainability into risk assessment were globally implemented and diffused. In this framework, the insurance sector is increasingly involved in addressing global environmental challenges even if it is traditionally seen as a relatively less impacting industry on climate issues. As climate change, the need for eco-innovation within the insurance industry has become more pressing. Regulatory frameworks worldwide are integrating climate-related risks and sustainability into business practices, and in particular within the financial and insurance sectors. In the European Union regulations such as Solvency II [European Insurance and Occupational Pensions Authority (EIOPA), 2009] − which includes climate risk integration − and the Sustainable Finance Disclosure Regulation (European Commission, 2021) − which requires transparency on sustainability impacts − demonstrate the growing emphasis on sustainable finance. The UK has implemented mandatory Task Force on Climate-related Financial Disclosures (Prudential Regulation Authority, 2021), and the USA has introduced the Climate Risk Disclosure Survey (National Association of Insurance Commissioners, 2023) together with New York State Department of Financial Services Climate Risk Guidance (New York State Department of Financial Services, 2022). In Asia, the Monetary Authority of Singapore − through the Green Insurance Framework (Monetary Authority of Singapore, 2022) and the China’s Banking and Insurance Regulatory Commission through the Green Insurance Guidelines (China Banking and Insurance Regulatory Commission, 2021) − promote sustainable insurance practices. In Australia, the Australian Prudential Regulation Authority through the Climate Risk Prudential Standard (Australian Prudential Regulation Authority, 2023) sets the standards for climate risk management. These regulatory developments, provided as example to support the idea underlying this paper, underscore the growing recognition of the insurance sector’s impact on sustainability. Simultaneously, the OECD highlights that eco-innovation possesses two key distinguishing features compared to conventional innovation: “Eco-innovation can, however, be distinguished from conventional innovation in two significant ways. First, it is not an open-ended concept as it represents innovation which explicitly emphasizes the reduction of environmental impacts, whether intended or not. Second, eco-innovation is not limited to innovation in products, processes, marketing methods and organisational methods, but also includes innovation in social and institutional structures [Organisation for Economic Co-operation and Development (Paris), 2009; p. 40]” [Organisation for Economic Co-operation and Development (OECD), 2010]. Then, eco-innovation, defined as the development of new products, processes or business models that reduce environmental impact while enhancing, at the same time, economic performance, is emerging as a strategy for insurers to manage risks, comply with regulatory requirements and respond to the growing demand for sustainable practices (Carrión-Flores and Innes, 2010; Dechezleprêtre et al., 2013). Within this evolving context, the insurance sector is increasingly called upon to address global environmental challenges − even though it has traditionally been considered a less impactful industry in terms of climate issues. However, as climate change escalates, the need for eco-innovation within insurance companies becomes more pressing. Surprisingly, although the key role that the financial sector can play in promoting and adopting green practices, to the best of our knowledge no study has so far analyzed the main governance determinants explaining insurance companies involvement in eco-innovation initiatives. Our paper tries to overcome this limit. Using two different regressions models, we test our research question by analyzing data from 252 insurance companies from 2011 to 2022.
We innovate previous literature from different points of view. First, we study for the first time the impact of governance toward a new eco-innovation index developed by Albitar et al. (2023). Second, we investigate this relationship within the insurance industry and consider a global sample. In this respect, studies on the green practices of insurance companies are quite limited, as the literature mainly focuses on nonfinancial companies (Sehnem et al., 2016)) or banks (Nath et al., 2014; Chen et al., 2022). Therefore, our analysis extends the previous literature. Finally, in our knowledge, our study is the first to test, through a survival analysis, the effect of adopting sustainability incentives, the presence of an environmental management team and corporate social responsibility (CSR) committees in achieving specific levels of eco-innovation.
The structure of the paper is as follows. Section 2 reviews the extant literature on eco-innovation and corporate governance. Section 3 outlines data and methods used in this study. Section 4 presents results and discussion results of our empirical analysis. Finally, Section 5 concludes.
2. Literature review
Among this framework, corporate governance, as a determinant of firms’ behavior, drives eco-innovation, and this is the idea underlying the analysis proposed in our paper. Governance mechanisms, such as board diversity, independence and the presence of sustainability committees, influence firms’ environmental strategies and performance (Ludwig and Sassen, 2022; Walls et al., 2012). For instance, diverse boards are more likely to consider a broader range of stakeholder interests like environmental concerns; on the other side, independent directors can provide the necessary independent advice to ensure that sustainability goals are integrated into corporate strategies (Ben-Amar et al., 2017; Cucari et al., 2018). Moreover, the establishment of board-level sustainability committees improve environmental performance and transparency: these committees, in fact, monitor and report different sustainability initiatives (Dixon-Fowler et al., 2017; Orazalin, 2020). In this context, the role of leadership is also important. Especially the figure of the chief executive officer (CEO) is fundamental in promoting eco-innovation. CEOs who are able to have a long-term perspective and to have a higher commitment to sustainability are more likely to promote environmental initiatives and allocate resources within eco-innovative projects (Ortiz-de-Mandojana et al., 2019). Another factor that can affect sustainability are the incentive structures: aligning executive compensation with environmental performance can further motivate CEOs to prioritize sustainability (Berrone and Gomez-Mejia, 2009). Additionally, ownership structures, such as state or public ownership, have been found to positively influence firms’ environmental proactivity, as these owners often have a vested interest in long-term sustainability (Calza et al., 2016; Lau et al., 2016). In other terms, corporate governance affects company’s sustainability. In fact, companies are increasingly recognizing the importance of integrating sustainability into their governance structures (Organisation for Economic Co-operation and Development, 2019). On this regard, companies therefore encouraged to voluntarily adopt transparent and responsible business practices to align with sustainability objectives (World Business Council for Sustainable Development, 2019). A detailed and original indicator to measure corporate commitment to climate change has been proposed in literature (Albitar et al., 2023). Examining the bidirectional relationship between environmental innovation and pollution reduction (Carrión-Flores and Innes, 2010) in the context of innovation, it has been demonstrated that innovation decreases toxic emissions and simultaneously contribute to further innovation. Furthermore, in the context of governance and its impact on eco-innovation, Amore and Bennedsen (2016) investigate the relationship between corporate governance and environmental innovation. Their study highlights that weak governance structures, particularly those with low institutional ownership and financial constraints, can obstruct environmental efficiency; the study suggests that effective governance raises sustainable advancements. Another relevant study analyzes the impact of China’s New Environmental Protection Law (NEPL) on corporate green innovation (Zhang et al., 2024). The study highlights that a positive management discussion and analysis tone amplifies the NEPL’s effect on innovation; despite this aspect, its influence on innovation quality is limited. Additionally, exploring the relationship between corporate governance and sustainability (Aras and Crowther, 2008), it is evident that both concepts are fundamental to corporate longevity. Their analysis of FTSE100 companies highlights both strengths and gaps in governance structures; in particular, it suggests that effective corporate governance can help to achieve sustainable development and long-term financial performance (but also that further analysis is required to define these relationships). To have a view in the field, a systematic literature review (E-Vahdati et al., 2019) on the integration of corporate governance and sustainability, identified vision, mission and leadership as drivers of sustainability. Relevant findings highlight regional variations in governance−sustainability integration and emphasize the need for further research to refine theoretical models and improve corporate strategies for sustainability. On this line, regarding corporate governance and green innovation, it has been demonstrated that stronger governance structures lead to increased investment in green R&D and improved environmental performance (Makpotche et al., 2024). Their findings show how effective governance mitigates agency conflicts, facilitates external financing and supports firms in addressing climate change through sustainable innovation. In the context of sustainable performance of financial intermediaries, environmental management and green product development remain undervalued (Kumar and Prakash, 2019). Paraschiv et al. (2012) explore some of the main factors related to corporate sustainability, and they emphasize how eco-innovation, responsible leadership and organizational change impact the integration of social and environmental aspects into corporate strategies. Salvioni and Gennari (2014) examine the relationship between corporate governance, sustainability and capital markets, demonstrating how sustainability-oriented governance bridges the gap between insider and outsider systems. They emphasize the importance of integrating sustainability into governance to enhance stakeholder engagement and long-term market-driven management. Pan et al. (2021) explore the relationship between CSR and eco-innovation from a triple bottom line perspective. Their study reveals that the social and financial dimensions of CSR have a positive moderating effect on sustainable environmental innovation. Regarding the role of board environmental committees in enhancing corporate environmental performance, findings demonstrate a positive association between the presence of such committees and improved environmental outcomes, particularly when supported by a senior-level sustainability manager (Dixon-Fowler et al., 2017). This suggests that governance mechanisms and managerial roles drive environmental performance. The study by Phung et al. (2023) examines the impact of top-management compensation on the environmental innovation strategy indicating that greater are the levels of top-management compensation, higher are the scores of eco-innovation engagement. Regarding the relationship between corporate governance and corporate sustainability, with a focus on the roles of boards of directors and investor relations officers, it has been demonstrated that internal governance forces, such as a higher proportion of inside directors, positively affect environmental and governance performance (Crifo et al., 2019). In addition, observing the relationship between corporate governance and sustainability disclosure, the board composition, leadership and structure are related to a company’s sustainability reporting practices (Michelon and Parbonetti, 2012). Three main thematic clusters on corporate governance and sustainability have been identified: CSR and reporting, corporate governance strategies and board composition (Naciti et al., 2022). This study highlights a shift from abstract concepts such as ethics and responsibility to more actionable themes, including board diversity and independent directors. Continuing the side of the empirical study on the relationship between innovation and corporate sustainability, it has been shown that service innovation performance significantly enhances intermediaries’ sustainability outcomes (Forcadell et al., 2019). Ben-Amar et al. (2017) demonstrate that gender-diverse boards are more effective in addressing stakeholder demands for transparency and sustainability, reinforcing global efforts to promote gender diversity in corporate governance. Corporate governance also affects CSR disclosure in banking sector: board gender diversity positively affects CSR reporting, while board size and independence have no significant impact (Orazalin, 2019). The board involvement in corporate sustainability reporting affects eco-practices. In particular, firms with larger boards, a higher proportion of independent directors and more female directors are more likely to adopt sustainability disclosure policies (Rathnayaka Mudiyanselage, 2018). In essence, the board composition affects sustainability transparency. Ferrero‐Ferrero et al. (2015) examine the mediating role of CSR management quality in the relationship between board diversity, particularly generational diversity, and CSR performance. They demonstrate that diverse boards enhance the design of sustainable strategies, and their findings highlight the importance of generational diversity in corporate governance as a driver to integrate sustainability into business practices. Indeed, Cucari et al. (2018) analyze the relationship between board diversity − including CSR committees and independent directors − and ESG disclosure in Italian listed companies, finding that independent directors and CSR committees are positively associated to ESG reporting. Their study provides insights on the role of board diversity in enhancing ESG transparency and governance practices. Kock et al. (2012) studied the impact of corporate governance mechanisms − including board composition, managerial incentives and regulatory systems − on firms’ environmental performance and demonstrated that governance structures is related to the adoption of greener corporate practices. Alipour et al. (2019) tested the relationship between environmental disclosure quality and firm performance, finding that board independence strengthens this relation: independent directors promote corporate oversight, reduce managerial self-interest and improve sustainability-related financial performance. Moreover, governance strongly influence environmental and ESG disclosure scores, with board independence and meeting attendance positively correlating with higher sustainability performance. In fact, “board meeting attendance is an important predictor of both scores, suggesting that more disciplined boards result in better sustainability performance” (Shrivastava and Addas, 2014). On the dimension of the board, Zaman et al. (2024) find that larger boards, as part of structural diversity, positively affect eco-innovation by providing a broader range of expertise and perspectives, which enhances strategic decision-making for environmental sustainability. In addition, board members with foreign experience and larger board sizes positively correlate with higher CSR performance (Lau et al., 2016). Then, diverse expertise and broader perspectives within boards are more effective in driving CSR initiatives. A positive role of board size and composition in driving sustainable performance, offering a practical tool for investors and banks to evaluate and improve ESG integration, is acknowledged (Porzio and Battaglia, 2024). Also having specific skills is important in the relationship between corporate governance and eco-innovation: Del Río et al. (2016) highlight that firms with strong industry-specific and financial expertise on their boards are better equipped to develop eco-innovations, as these specialized skills enable more effective resource allocation and strategic decision-making for sustainability initiatives. Together with these aspects, it is relevant to note that CEO duality impact CSR disclosure, particularly in social and governance dimensions. This is a dynamic that warrants further investigation to understand its implications for eco-innovation adoption, where strategic decisions often require long-term commitment and stakeholder alignment (Lassoued and Khanchel, 2023). Huang (2013) investigates the influence of CEO demographic characteristics, such as educational background (MBA and MSc), tenure and gender, on CSR performance. Data reveal a significant association between these factors and the consistency of firms’ CSR rankings. The study underscores the importance of integrating CSR into academic curricula and highlights the role of leadership traits in driving sustainable corporate development. One of the most important studies in this field by Agnese et al. (2023) empirically demonstrate that board gender diversity and the presence of a CSR sustainability committee significantly enhance banks’ environmental performance, particularly in emissions reduction, environmental innovation and resource efficiency. Additionally, firms with sustainability committees implement more effective CSR strategies, which in turn drive improved sustainability outcomes, reinforcing the importance of governance structures in corporate sustainability efforts (Orazalin, 2020). Furthermore, the study by Allodi and Soana (2024) demonstrates that environmental strategies and sustainability committees significantly promote circular economy outcomes in banks, while board gender diversity is positively related to emission reduction and green products offered, but not to e-waste reduction. Dixon-Fowler et al. (2017) conducted a study on the role of board environmental committees in enhancing corporate environmental performance, finding a positive association between the presence of such committees and improved environmental outcomes, particularly when supported by a senior-level sustainability manager. They suggest that governance mechanisms and managerial roles drive environmental performance. Ultimately, it has been demonstrated that environmental management capabilities have a significant positive relation on circular eco-innovation (Scarpellini et al., 2020). Given these findings, it is relevant to consider the role of a dedicated environment management team in eco-innovation, as prior studies have highlighted how managerial environmental awareness support green innovation (Peng and Liu, 2016). To conclude, our research question investigates whether and how corporate governance is related to the eco-innovation of insurance companies.
3. Data and methods
Our sample is constructed using data from Thomson Reuters Refinitiv Eikon database. We select all insurance-listed companies operating during the year 2022. To improve the robustness of our analysis, we consider a relatively long time span and retrieve economic and ESG data for the previously selected companies over the period 2011–2022, both operating in developed (154 insurance companies) and developing countries (98 insurance companies). Following Giráldez-Puig et al. (2025), we chose 2011 as the starting point because that year marked the beginning of the United Nations’ efforts to raise awareness about the upcoming Rio + 20 Conference in June 2012, emphasizing the importance of incorporating ESG considerations into the management of insurance companies. Moreover, for many firms, data prior to 2011 were not available. Our final sample consists of 252 insurance companies across various geographical areas, resulting in a total of 1,434 observations, collected annually over the period 2011–2022. Table 1 reports the distribution of these 252 companies by mainland and business model, providing a clearer picture of the sample composition. The largest share of companies operates in America (39.7%), followed by Asia (32.9%) and Europe (20.6%). In terms of business model, property and casualty insurance accounts for 34.5% of the sample, multiline insurance and brokers for 24.2% and reinsurance for 5.2%, while 36.1% of firms fall into the “Not specified” category. Africa and Oceania are represented by smaller shares, accounting for 3.6% and 3.2% of the sample, respectively.
Sample composition by mainland and business model
| Mainland | Not specified | Multiline insurance and brokers | Property and casualty insurance | Re- insurance | Total | Total (%) |
|---|---|---|---|---|---|---|
| Africa | 6 | 2 | 1 | 0 | 9 | 3.6 |
| America | 27 | 17 | 54 | 2 | 100 | 39.7 |
| Asia | 39 | 20 | 17 | 7 | 83 | 32.9 |
| Europe | 17 | 20 | 11 | 4 | 52 | 20.6 |
| Oceania | 2 | 2 | 4 | 0 | 8 | 3.2 |
| Total | 91 | 61 | 87 | 13 | 252 | 100 |
| Mainland | Not specified | Multiline insurance and brokers | Property and casualty insurance | Re- insurance | Total | Total (%) |
|---|---|---|---|---|---|---|
| Africa | 6 | 2 | 1 | 0 | 9 | 3.6 |
| America | 27 | 17 | 54 | 2 | 100 | 39.7 |
| Asia | 39 | 20 | 17 | 7 | 83 | 32.9 |
| Europe | 17 | 20 | 11 | 4 | 52 | 20.6 |
| Oceania | 2 | 2 | 4 | 0 | 8 | 3.2 |
| Total | 91 | 61 | 87 | 13 | 252 | 100 |
This table reports the sample distribution by mainland and business model for the whole sample period (2011–2022)
We measure the eco-innovation score, our dependent variable, through the eco-innovation index (ECOINN) developed by Albitar et al. (2023) and constructed considering five eco-innovation proxies that reflect companies’ efforts to reduce the impact on the environment. Specifically, these authors identify in the Refinitv Eikon database the following five environmental indicators:
eco-design products;
environmental assets under management;
environmental products;
product environmental responsible use; and
renewable/clean energy products.
The indicators take value 1 if the firm adopts each environmental friendly practices, and 0 otherwise. Therefore, the eco-innovation index (ECO) ranges from 0 (lowest level of eco-innovation) to 5 (highest level of eco-innovation). Table 2 reports the definitions of the five indicators described above.
Eco-innovation indicator definitions
| CE indicators | Description |
|---|---|
| (i) Eco-design products | A dummy variable that equals 1 if the company reports on specific products which are designed for reuse, recycling or the reduction of environmental impacts, 0 otherwise |
| (ii) Environmental assets under management | A dummy variable that equals 1 if the company reports on assets under management which employ environmental screening criteria or environmental factors in the investment selection process, 0 otherwise |
| (iii) Environmental products | A dummy variable that equals 1 if the company reports on at least one product line or service that is designed to have positive effects on the environment or which is environmentally labeled and marketed, 0 otherwise |
| (iv) Product environmental responsible use | A dummy variable that equals 1 if the company reports about product features and applications or services that will promote responsible, efficient, cost-effective and environmentally preferable use, 0 otherwise |
| (v) Renewable/clean energy products | A dummy variable that equals 1 if the company develops products or technologies for use in the clean, renewable energy (such as wind, solar, hydro, geo-thermal and biomass power), 0 otherwise |
| Description | |
|---|---|
| (i) Eco-design products | A dummy variable that equals 1 if the company reports on specific products which are designed for reuse, recycling or the reduction of environmental impacts, 0 otherwise |
| (ii) Environmental assets under management | A dummy variable that equals 1 if the company reports on assets under management which employ environmental screening criteria or environmental factors in the investment selection process, 0 otherwise |
| (iii) Environmental products | A dummy variable that equals 1 if the company reports on at least one product line or service that is designed to have positive effects on the environment or which is environmentally labeled and marketed, 0 otherwise |
| (iv) Product environmental responsible use | A dummy variable that equals 1 if the company reports about product features and applications or services that will promote responsible, efficient, cost-effective and environmentally preferable use, 0 otherwise |
| (v) Renewable/clean energy products | A dummy variable that equals 1 if the company develops products or technologies for use in the clean, renewable energy (such as wind, solar, hydro, geo-thermal and biomass power), 0 otherwise |
This table reports the definitions of the indicators constituting the eco-innovation index
Although this index is originally designed to be broadly applicable across sectors, these indicators can also be meaningfully applied to insurance companies, with interpretations tailored to the nature of their business. For instance, while insurers do not produce and sell tangible goods, they can still engage in eco-design by developing insurance products and services that incorporate environmental considerations. This may include offering policies that provide incentives for electric vehicle ownership, support energy-efficient buildings or reduce paper usage through digital claims processing − all reflecting a thoughtful design that promotes environmental responsibility.
Similarly, the indicator relating to environmental assets under management is particularly relevant for insurance companies due to their role as institutional investors. When insurers integrate environmental criteria into their investment portfolios − such as through the inclusion of green bonds, ESG-compliant funds or direct investments in sustainable infrastructure − they actively contribute to the financing of the green transition, thereby enhancing their eco-innovation profile. Looking at the third indicator, insurance firms can also demonstrate eco-innovation by offering products with an explicit environmental focus. Furthermore, insurers can promote responsible environmental use through the features of their policies. This may involve encouraging customers to adopt sustainable behaviors, such as offering premium discounts for low-emission vehicles or for the use of eco-friendly materials in property repairs. Insurers may also enhance their environmental performance through the adoption of digital platforms that reduce the need for physical resources, contributing to operational sustainability. Finally, insurance companies can support renewable and clean energy by underwriting and facilitating the development of renewable energy projects. By providing customed insurance solutions for solar panels, wind farms or geothermal systems, insurers help manage the unique risks tied to clean energy projects, playing a key part in supporting and accelerating the sector’s growth.
Our independent variables are nine governance variables suggested by previous literature:
board gender diversity (Horbach and Jacob, 2018; Issa and Bensalem, 2023);
board meeting attendance (Zaman et al., 2024);
board size (Zaman et al., 2024; Roy and DasGupta, 2025);
board-specific skills (García‐Sánchez et al., 2021; Ullah et al., 2024);
CEO−chairman separation (Zaman et al., 2024; Owolabi et al., 2025);
CSR sustainability committee (Karaman et al., 2024; Alodat et al., 2025);
environment management team (Peng and Liu, 2016);
independent board members (García‐Sánchez et al., 2021; Zaman et al., 2024); and
sustainability compensation incentives (Phung et al., 2023; Daniel Vasconcelos and de Ribeiro, 2025).
Board gender diversity (GEND) is estimated by the percentage of female directors on the board of directors. Board meeting attendance (BMEET) and board size (BSIZE) represent the number of board meetings during the year and the number of board members. Board-specific skills (SKILLS) is computed as the percentage of board members who have either an industry-specific background or a strong financial background. CEO−chairman separation (CEO) is measured through a dummy variable that takes value 1 if the CEO simultaneously chair the board or has the chairman of the board been the CEO of the company, 0 otherwise. CSR committee (CSR) and environment management team (EMT) are both measured by a dummy variable that takes value 1 if the company has a CSR committee/environment management team, 0 otherwise. Independent board members (INDEP) are estimated by the percentage of independent board members on the board. Finally, sustainability compensation incentives (SUST_INC) are measured by a dummy variable that takes value 1 if the senior executive’s compensation is linked to CSR/H&S/sustainability targets.
In addition, we consider some specific insurance variables as controls: (i) size (SIZE), represented by the natural logarithm of the total premium written by the insurance company; (ii) profitability, estimated by the return on assets (ROA), calculated by dividing net income by total assets; (iii) leverage (LEV), calculated as the ratio of total debts to total assets; (iv) a measure of risk, estimates by the loss ratio (LOSS), which denotes the ratio of losses to premiums earned, encompassing paid insurance claims and adjustment expenses. Table 3 reports a comprehensive presentation of all the research variables described above.
Variable measurement description
| Variables | Symbols | Description |
|---|---|---|
| Dependent variable | ||
| Eco-innovation index | ECOINN | Score of the firm adoption of eco-innovation practices. The score ranges from 0 (no practices) to 5 (all practices) |
| Independent variables | ||
| Board gender diversity | GEND | Percentage of female on the board of directors |
| Board meeting attendance | BMEET | The number of board meetings during the year |
| Board size | BSIZE | Total number of board members |
| Board-specific skills | SKILLS | Percentage of board members who have either an industry-specific background or a strong financial background |
| CEO-chairman separation | CEO | Dummy variable that takes value 1 if the CEO simultaneously chair the board or has the chairman of the board been the CEO of the company, 0 otherwise |
| CSR sustainability committee | CSR | Dummy variable that takes value 1 if the company has a corporate social responsibility committee, 0 otherwise |
| Environment management team | EMT | Dummy variable that takes value 1 if the company has an Environment Management Team, 0 otherwise |
| Independent board members | INDEP | Percentage of independent board members |
| Sustainability compensation incentives | SUST_INC | Dummy variable that takes value 1 if the senior executive’s compensation is linked to CSR/H&S/Sustainability targets |
| Control variables | ||
| Size | SIZE | The natural logarithm of the total premium written by the insurance company |
| Return on assets | ROA | The ratio between net income by total assets |
| Leverage | LEV | The ratio of total debts to total assets |
| Loss ratio | LOSS | The ratio of losses to premiums earned |
| Variables | Symbols | Description |
|---|---|---|
| Dependent variable | ||
| Eco-innovation index | Score of the firm adoption of eco-innovation practices. The score ranges from 0 (no practices) to 5 (all practices) | |
| Independent variables | ||
| Board gender diversity | Percentage of female on the board of directors | |
| Board meeting attendance | The number of board meetings during the year | |
| Board size | Total number of board members | |
| Board-specific skills | Percentage of board members who have either an industry-specific background or a strong financial background | |
| CEO-chairman separation | Dummy variable that takes value 1 if the | |
| Dummy variable that takes value 1 if the company has a corporate social responsibility committee, 0 otherwise | ||
| Environment management team | Dummy variable that takes value 1 if the company has an Environment Management Team, 0 otherwise | |
| Independent board members | Percentage of independent board members | |
| Sustainability compensation incentives | SUST_INC | Dummy variable that takes value 1 if the senior executive’s compensation is linked to CSR/H&S/Sustainability targets |
| Control variables | ||
| Size | The natural logarithm of the total premium written by the insurance company | |
| Return on assets | The ratio between net income by total assets | |
| Leverage | The ratio of total debts to total assets | |
| Loss ratio | The ratio of losses to premiums earned | |
This table describes the variables and their definitions
3.1 Research methodology
To test our research question, we run the following regression model:
where Y is the insurance’s eco-innovation index, GOVi,t and CONTROLSi,t are vectors considering respectively the governance and controls variables described above. Finally, Ɛ represents the error term which is assumed to be independent and identically distributed.
We consider three estimation approaches − pooled ordinary least squares (pooled OLS), fixed effects (FE) and random effects (RE) − to determine the most appropriate specification for our data. The fixed effects model accounts for unobserved heterogeneity across firms by allowing individual-specific intercepts, while the random effects model assumes that such heterogeneity is uncorrelated with the explanatory variables.
To determine the most appropriate specification, we perform a Hausman test, which evaluates whether the unique errors are correlated with the regressors. The results of the test do not reject the null hypothesis, suggesting that the random effects model is preferable to the fixed effects model. Therefore, in the analysis we report only the results of the pooled OLS and random effects estimations. The inclusion of the pooled OLS model serves as a benchmark for comparison, while the random effects model is retained due to its efficiency under the assumptions supported by the Hausman test. This dual approach allows us to robustly test our research question while accounting for both baseline estimations and panel-specific effects.
Our findings are also subjected to robustness checks (Section 4.3) to assess their stability under alternative specifications. Specifically, we test the robustness of our results by: (i) using an alternative dependent variable to the eco-innovation Index; (ii) exploring differences between developed and developing countries; and (iii) including a country-level environmental control variable. We chose not to incorporate these country-level variables (e.g. Environmental Performance Index and the Developed dummy) in the baseline regressions because our primary focus is on firm-level governance determinants rather than institutional or macroeconomic factors. Including them in the main specification could have shifted the emphasis of the analysis and reduced the clarity of our contribution. Furthermore, the inclusion of these variables would have considerably reduced the sample size, due to data availability, and potentially introduced multicollinearity with firm-specific controls. For these reasons, we opted to introduce them only in the robustness checks to verify that our main findings hold even after accounting for broader contextual differences, thus reinforcing the validity of our conclusions without compromising model parsimony.
Finally, in addition to our main regression analysis, we conduct a survival analysis as an additional method to enrich our interpretation of the results. Specifically, we apply the Kaplan−Meier estimator to examine how the probability of achieving a high level of eco-innovation evolves over time across firms with different governance features (e.g. the presence of an environmental management team, a CSR sustainability committee or sustainability compensation incentives).
We also estimate a Cox proportional hazards model to assess the influence of these governance variables on the likelihood of becoming eco-innovative. In this context, the dependent variable is a binary indicator that takes the value of 1 when a firm achieves and maintains an eco-innovation score greater than 2, and 0 otherwise. The time variable corresponds to the number of years until the firm reaches this eco-innovation threshold. This analysis allows us to capture not only whether firms become eco-innovative, but also how quickly they do so, based on different governance structures.
4. Results and discussion
4.1 Descriptive statistics
Table 4 provides descriptive statistics for the variables included in the analysis. Our dependent variable, the eco-innovation index − calculated for each of the 1,434 firm-year observations, based on annual data for each firm over the period 2011–2022 − reveals a relatively low mean value of 1.388 (SD = 1.634) on a scale ranging from 0 to 5, indicating limited engagement in eco-innovative practices across the sample. Figure 1 shows the eco-innovation index for each country. Board gender diversity (GEND) has a mean value of 21.03%, indicating a moderate level of female representation. The standard deviation of 13.34% and a range from 0% to 71.43% reveal substantial variation across firms. While some companies have embraced gender diversity, others still lack female board members entirely. Board meeting attendance (BMEET) reports a high average attendance rate of 89.01% (SD = 11.35), indicating a generally strong level of board engagement. Board size (BSIZE) shows a mean of 11.08 members (SD = 3.39), with board sizes varying from 1 to 41 members, which is consistent with governance best practices recommending mid-sized boards to balance diversity and effectiveness.
Descriptive statistics
| Variable | Obs | Mean | SD | Min. | Max. |
|---|---|---|---|---|---|
| Eco-innovation index | 1,434 | 1.388 | 1.634 | 0 | 5 |
| Board gender diversity | 1,434 | 21.029 | 13.335 | 0 | 71.430 |
| Board meeting attendance | 1,434 | 89.007 | 11.349 | 4 | 100 |
| Board size | 1,434 | 11.082 | 3.388 | 3 | 38 |
| Board-specific skills | 1,434 | 51.600 | 21.227 | 0 | 100 |
| CEO chairman separation | 1,434 | 0.353 | 0.478 | 0 | 1 |
| CSR sustainability committee | 1,434 | 0.543 | 0.498 | 0 | 1 |
| Environment management team | 1,434 | 0.361 | 0.480 | 0 | 1 |
| Independent board members | 1,434 | 68.746 | 21.368 | 0 | 100 |
| Sustainability compensation incentives | 1,434 | 0.219 | 0.414 | 0 | 1 |
| Size | 1,434 | 2.253 | 2.444 | 7.524 | 10.016 |
| ROA | 1,434 | 2.028 | 3.817 | −43.580 | 56.940 |
| Leverage | 1,434 | 6.392 | 7.698 | 0 | 98.844 |
| Loss ratio | 1,434 | 124.562 | 514.460 | 1.710 | 13,793.780 |
| Variable | Obs | Mean | Min. | Max. | |
|---|---|---|---|---|---|
| Eco-innovation index | 1,434 | 1.388 | 1.634 | 0 | 5 |
| Board gender diversity | 1,434 | 21.029 | 13.335 | 0 | 71.430 |
| Board meeting attendance | 1,434 | 89.007 | 11.349 | 4 | 100 |
| Board size | 1,434 | 11.082 | 3.388 | 3 | 38 |
| Board-specific skills | 1,434 | 51.600 | 21.227 | 0 | 100 |
| 1,434 | 0.353 | 0.478 | 0 | 1 | |
| 1,434 | 0.543 | 0.498 | 0 | 1 | |
| Environment management team | 1,434 | 0.361 | 0.480 | 0 | 1 |
| Independent board members | 1,434 | 68.746 | 21.368 | 0 | 100 |
| Sustainability compensation incentives | 1,434 | 0.219 | 0.414 | 0 | 1 |
| Size | 1,434 | 2.253 | 2.444 | 7.524 | 10.016 |
| 1,434 | 2.028 | 3.817 | −43.580 | 56.940 | |
| Leverage | 1,434 | 6.392 | 7.698 | 0 | 98.844 |
| Loss ratio | 1,434 | 124.562 | 514.460 | 1.710 | 13,793.780 |
This table shows the summary statistics of the variables included in equations (1) in the period 2011–2022. Variable definitions are provided in Table 3
The map displays global scores across countries using a green gradient scale, where darker shades represent higher scores. Countries with notable emphasis include the United States, Brazil, India, and China, which show stronger values compared to lighter-shaded regions such as Africa and parts of Europe. The shading indicates the distribution of scores across continents, providing a comparative geographical overview.Global map of eco-innovation index by country
Note(s): This figure shows the average of eco-innovation index for each country in the period 2011–2022
Source: Authors’ own creation
The map displays global scores across countries using a green gradient scale, where darker shades represent higher scores. Countries with notable emphasis include the United States, Brazil, India, and China, which show stronger values compared to lighter-shaded regions such as Africa and parts of Europe. The shading indicates the distribution of scores across continents, providing a comparative geographical overview.Global map of eco-innovation index by country
Note(s): This figure shows the average of eco-innovation index for each country in the period 2011–2022
Source: Authors’ own creation
Regarding board-specific skills (SKILLS), the average proportion of directors with relevant skills is 51.60% (SD = 21.23) suggesting that, on average, about half of board members possess relevant expertise, though the broad range (0%–100%) reveals significant variability in skill composition. The binary variable CEO−chairman separation (CEO) indicates that 35.3% of the firms in the sample separate the roles of CEO and chairman. Similarly, CSR sustainability committee (CSR) is present in 54.3% of the firms, while environmental management teams (EMT) are observed in 36.1% of firms suggesting that while formal sustainability oversight is becoming common, operational environmental integration remains less widespread.
Independent board members (INDEP) constitute, on average, 68.75% of the board (SD = 21.37), with values ranging from 0% to 100%. Sustainability compensation incentives, another binary variable, are adopted by 21.9% of the firms highlighting that tying executive pay to sustainability performance is still an emerging practice. Among control variables the size (SIZE) reports a mean of 2.253 (SD = 2.444), while return on assets (ROA), shows a relatively low average of 2.03 (SD = 3.817), suggesting wide variation in profitability across the sample. Leverage shows a mean of 6.392 (SD = 7.698), reflecting considerable variability in firms’ use of debt. This suggests a mix of conservative and more aggressively financed firms within the sample. The loss ratio also presents a highly skewed distribution, with a mean of 124.562 and a maximum exceeding 13,000.
4.2 Baseline results and discussion
The regression results, reported in Table 5, shed light on the drivers of eco-innovation, emphasizing the roles of board composition, governance structures and firm-level characteristics.
Regressions on eco-innovation index
| Variables | (1) | (2) |
|---|---|---|
| GEND | 0.015*** (0.003) | 0.017*** (0.003) |
| BMEET | 0.015*** (0.003) | 0.011*** (0.003) |
| BSIZE | 0.007 (0.010) | 0.014 (0.012) |
| SKILLS | −0.008*** (0.001) | −0.006*** (0.001) |
| CEO | 0.073 (0.066) | 0.031 (0.085) |
| CSR | 0.974*** (0.078) | 0.800*** (0.070) |
| EMT | 0.548*** (0.074) | 0.505*** (0.074) |
| INDEP | 0.005*** (0.002) | 0.006*** (0.002) |
| SUST_INC | 0.341*** (0.078) | 0.244*** (0.063) |
| ROA | −0.037*** (0.008) | −0.002 (0.008) |
| LOSS | 0.0001** (0.0001) | 0.0001 (0.0001) |
| SIZE | 0.191*** (0.016) | 0.192*** (0.023) |
| LEV | 0.004 (0.004) | −0.00004 (0.006) |
| Constant | −1.458*** (0.347) | −1.340*** (0.370) |
| Observations | 1,434 | 1,434 |
| R-squared | 0.515 | 0.331 |
| Fixed effects | No | No |
| Pooled OLS | Yes | No |
| Random effects | No | Yes |
| Variables | (1) | (2) |
|---|---|---|
| 0.015 | 0.017 | |
| 0.015 | 0.011 | |
| 0.007 (0.010) | 0.014 (0.012) | |
| −0.008 | −0.006 | |
| 0.073 (0.066) | 0.031 (0.085) | |
| 0.974 | 0.800 | |
| 0.548 | 0.505 | |
| 0.005 | 0.006 | |
| SUST_INC | 0.341 | 0.244 |
| −0.037 | −0.002 (0.008) | |
| 0.0001 | 0.0001 (0.0001) | |
| 0.191 | 0.192 | |
| 0.004 (0.004) | −0.00004 (0.006) | |
| Constant | −1.458 | −1.340 |
| Observations | 1,434 | 1,434 |
| R-squared | 0.515 | 0.331 |
| Fixed effects | No | No |
| Pooled | Yes | No |
| Random effects | No | Yes |
This table provides the estimates of the pooled OLS and random-effect regressions for 2011–2022. The dependent variable is the eco-innovation index (ECOINN). Variables definitions are provided in Table 3. Robust standard errors (SE) are reported in parentheses. ***p < 0.01; **p < 0.05 and *p < 0.10
Board gender diversity (GEND) shows a highly significant positive relationship with eco-innovation index in pooled ordinary least square model − Column (1) − and random effect model − Column (2).
This is in line with previous literature that found similar results on financial intermediaries and environmental performance (García-Sánchez, et al., 2018; Birindelli, et al., 2019; Galletta et al., 2022; Agnese et al., 2023) and insurance companies (Allodi et al., 2025). This finding may be interpreted by acknowledging that female board members often possess psychological characteristics such as empathy, sensitivity and a strong concern for others’ well-being. These traits contribute to a heightened awareness of environmental risks and a greater sense of environmental responsibility compared to their male counterparts (Burkhardt et al., 2020). As a result, women tend to be more actively involved in environmentally responsible behaviors (Hur et al., 2016).
Board meeting attendance (BMEET) is positively significant − Columns (1) and (2) − indicating that higher board engagement facilitates the oversight and strategic alignment required for eco-innovation. Board size (BSIZE) does not exhibit consistent significance, suggesting that the sheer number of board members is not linked to our dependent variable. Board-specific skills (SKILLS) consistently show a significant negative relationship with eco-innovation across all models. This finding aligns with the expectation that financially skilled boards may prioritize short-term profitability and risk management over long-term investments in sustainability, particularly in sectors where environmental innovation does not immediately align with financial performance. The presence of a CSR sustainability committee (CSR) has a significant positive coefficient in all models.
This finding underscores the central role of dedicated sustainability structures in embedding environmental objectives within corporate strategy, thereby promoting a more structured and proactive approach to eco-innovation. Moreover, these results are in line with the resource dependency theory, which posits that sustainability committees at the board level enhance organizational commitment to environmental goals, facilitate the effective implementation of CSR practices and support the formulation of sustainability-oriented strategies (Hussain et al., 2018).
Similarly, the environmental management team (EMT) is related to eco-innovation in both models − Columns (1) and (2) − suggesting that operational-level focus on sustainability supports the implementation of innovative practices. This finding is particularly relevant for insurance companies, where environmental sustainability increasingly intersects with risk management and product innovation. Prior studies have shown that firms can enhance their environmental outcomes by creating green teams among employees (Dangelico and Pontrandolfo, 2015; Hanna et al., 2000) and, more broadly, by implementing green human resource management systems (Dumont et al., 2017; Majid et al., 2020; Zhang et al., 2019). For insurance companies, these practices help leverage internal expertise not only to improve environmental efficiency within the organization, but also to better identify sustainability-related risks and to design innovative, eco-oriented insurance solutions (Daily et al., 2012; Antonioli et al., 2013).
The role of independent board members (INDEP) emerges as positively significant in all the models. This aligns with the perspective that independent directors enhance governance by challenging conventional strategies and supporting innovation initiatives, particularly those with long-term benefits such as environmental innovation. Sustainability compensation incentives (SUST_INC) are positively significant across all models, reflecting the alignment of executive incentives with eco-innovation goals. This finding underscores that performance-based compensation tied to sustainability objectives is an effective mechanism to promote environmental initiatives. Compensation mechanisms that take environmental issues into account can motivate managers to improve their attitude toward green initiatives and leverage their knowledge to address environmental challenges (Zhang et al., 2019), thereby improving their ability to nurture sustainability performance (Luzzini et al., 2015).
Among control variables, return on assets (ROA) shows a significant negative association with eco-innovation only considering pooled ols model − Column (1). This suggests a potential conflict between achieving high short-term profitability and committing resources to sustainability, as eco-innovation often requires upfront investments with deferred returns. Also the loss ratio (LOSS) is marginally significant only referring to pooled OLS model − Column (1) − suggesting that higher claims might pressure insurance firms to adopt innovative practices to improve operational efficiency. However, the weak significance indicates this relationship may vary across contexts. Firm Size (SIZE) positively correlates with green practices − Columns (1) and (2) − reflecting that larger firms, with more substantial resources and reputational concerns, are better positioned to invest in sustainable innovations.
4.3 Robustness tests
We test the robustness of our findings using several tests. First, we test our models on a further dependent variable related to eco-innovation: environmental innovation score (EIS). This variable, extracted from Refinitiv Eikon, reflects the company’s capacity to reduce the environmental costs and burdens for its customers, thereby creating new market opportunities through new environmental technologies and processes or eco-designed products. Being a score it is presented as a percentage, ranging between 0% and 100%. The results are reported in Table 6.
Regressions on environmental innovation score
| Variables | (1) | (2) |
|---|---|---|
| GEND | 0.267*** (0.058) | 0.237*** (0.056) |
| BMEET | 0.239*** (0.066) | 0.166*** (0.064) |
| BSIZE | 0.229 (0.212) | 0.645*** (0.249) |
| SKILLS | −0.184*** (0.032) | −0.147*** (0.030) |
| CEO | 1.250 (1.428) | 4.693*** (1.733) |
| CSR | 17.259*** (1.680) | 13.161*** (1.402) |
| EMT | 10.521*** (1.601) | 10.083*** (1.503) |
| INDEP | 0.078** (0.036) | 0.129*** (0.045) |
| SUST_INC | 7.685*** (1.684) | 7.214*** (1.258) |
| Controls | Yes | Yes |
| Constant | −23.555*** (7.471) | −25.881*** (7.541) |
| Observations | 1,434 | 1,434 |
| R-squared | 0.452 | 0.278 |
| Fixed effects | No | No |
| Pooled OLS | Yes | No |
| Random effects | No | Yes |
| Variables | (1) | (2) |
|---|---|---|
| 0.267 | 0.237 | |
| 0.239 | 0.166 | |
| 0.229 (0.212) | 0.645 | |
| −0.184 | −0.147 | |
| 1.250 (1.428) | 4.693 | |
| 17.259 | 13.161 | |
| 10.521 | 10.083 | |
| 0.078 | 0.129 | |
| SUST_INC | 7.685 | 7.214 |
| Controls | Yes | Yes |
| Constant | −23.555 | −25.881 |
| Observations | 1,434 | 1,434 |
| R-squared | 0.452 | 0.278 |
| Fixed effects | No | No |
| Pooled | Yes | No |
| Random effects | No | Yes |
This table provides the estimates of the pooled OLS and random-effect regressions for 2011–2022. The dependent variable of interest is the environmental innovation score which reflects a company’s ability to reduce environmental costs and burdens for its customers while creating new market opportunities through the development of environmental technologies, processes or eco-designed products. Variables definitions are provided in Table 3. Robust standard errors (SE) are reported in parentheses. ***p < 0.01; **p < 0.05 and *p < 0.10
The tests confirm our main results. Specifically, regardless of the statistical methodology used, the presence of a CSR committee, the establishment of an environmental management team, the adoption of sustainability incentives and a higher proportion of independent directors have a significant and positive impact on the environmental innovation score. By contrast, an increased number of directors with high financial expertise leads to a decrease in the adoption of green practices by insurance companies. These findings are reasonable, as the establishment of CSR committees, sustainability incentives and the presence of an environmental management team are inherently “green-first” actions closely tied to sustainability. Conversely, the financial expertise of directors shows a negative relationship with the dependent variable, as it is reasonable to expect a stronger orientation toward financial performance rather than sustainability.
Second, we examine if the relationship changes between developed and developing countries by adding the variable “developed” to our regression. This is a dummy variable equal to 1 if the company’s headquarters is in a developed country, and 0 otherwise. The positive and significant effect of the mentioned dummy confirms that the institutional and economic context plays an important role in shaping company green behavior and outcomes. Results are reported in Table 7.
Regressions results including the “developed” dummy variable
| Variables | (1) | (2) |
|---|---|---|
| DEVELOPED | 0.443*** (0.078) | 0.324** (0.136) |
| GEND | 0.011*** (−0.003) | 0.016*** (0.003) |
| BMEET | 0.016*** (0.003) | 0.011*** (0.003) |
| BSIZE | 0.004 (0.010) | 0.01 (0.012) |
| SKILLS | −0.008*** (0.001) | −0.006*** (0.001) |
| CEO | −0.003 (0.067) | −0.003 (0.085) |
| CSR | 0.961*** (0.077) | 0.802*** (0.070) |
| EMT | 0.534*** (0.074) | 0.501*** (0.074) |
| INDEP | 0.002 (0.002) | 0.005** (0.002) |
| SUST_INC | 0.327*** (0.077) | 0.240*** (0.063) |
| Controls | Yes | Yes |
| Constant | −1.549*** (0.343) | −1.443*** (0.371) |
| Observations | 1.434 | 1.434 |
| R-squared | 0.526 | 0.335 |
| Pooled OLS | Yes | No |
| Random effects | No | Yes |
| Variables | (1) | (2) |
|---|---|---|
| DEVELOPED | 0.443 | 0.324 |
| 0.011 | 0.016 | |
| 0.016 | 0.011 | |
| 0.004 (0.010) | 0.01 (0.012) | |
| −0.008 | −0.006 | |
| −0.003 (0.067) | −0.003 (0.085) | |
| 0.961 | 0.802 | |
| 0.534 | 0.501 | |
| 0.002 (0.002) | 0.005 | |
| SUST_INC | 0.327 | 0.240 |
| Controls | Yes | Yes |
| Constant | −1.549 | −1.443 |
| Observations | 1.434 | 1.434 |
| R-squared | 0.526 | 0.335 |
| Pooled | Yes | No |
| Random effects | No | Yes |
This table provides the estimates of the pooled OLS and random-effect regressions for 2011–2022. The dependent variable is the eco-innovation index (ECOINN), while the independent variable of interest is developed (DEVELOPED), a dummy variable equal to 1 if the company’s headquarters is in a developed country, and 0 otherwise. Variables definitions are provided in Table 3. Robust standard errors (SE) are reported in parentheses. ***p < 0.01; **p < 0.05 and *p < 0.10
Then, to explore our model in more depth, we split our sample between insurance companies from developed countries (154) and developing countries (98) to investigate more in detailed possible differences between the two subsamples. The results are reported in Table 8. In developed countries (Panel A), board gender diversity (GEND) is positively significant − Columns (1) and (2) − suggesting that greater gender diversity enhances eco-innovation. This finding aligns with literature (Allodi et al., 2025) emphasizing diverse boards fostering innovation and sustainability. However, in developing countries (Panel B), GEND is not significant. This discrepancy could be attributed to contextual differences; in developing nations, gender diversity might not yet translate into meaningful influence on board decisions, possibly due to socioeconomic and institutional constraints limiting the effectiveness of diverse perspectives. Board meeting attendance (BMEET) is strongly significant and positive in developed countries, reinforcing the importance of active governance in driving environmental innovation. However, it is not significant in developing countries (Panel B), indicating that formal attendance might not adequately capture board engagement or effectiveness in these contexts. The CSR sustainability committee (CSR) and environmental management team (EMT) are consistently significant and positively associated with the eco-innovation index across all models in both panels. These findings underscore the critical role of dedicated sustainability-oriented governance structures in fostering environmental innovation, regardless of economic development levels. Interestingly, board-specific skills (SKILLS) exhibit a negative relationship with eco-innovation in developed countries (Panel A), suggesting that financially skilled boards may prioritize short-term performance over long-term sustainability. This pattern is less pronounced in developing countries, where SKILLS are significant only with pooled OLS model (Panel B − Column 1). Finally, sustainability compensation incentives (SUST_INC) demonstrate a strong positive impact only in developed countries (Panel A), showing the greater institutional maturity and stronger alignment of incentive structures with environmental goals in developed economies.
Regressions on eco-innovation index in developed and developing countries
| PANEL A | PANEL B | |||
|---|---|---|---|---|
| Variables | (1) | (2) | (1) | (2) |
| GEND | 0.015*** (0.003) | 0.020*** (0.003) | −0.001 (0.006) | 0.005 (0.005) |
| BMEET | 0.024*** (0.003) | 0.020*** (0.004) | −0.008 (0.006) | −0.005 (0.005) |
| BSIZE | −0.007 (0.012) | 0.029* (0.015) | −0.002 (0.017) | −0.025 (0.021) |
| SKILLS | −0.009*** (0.002) | −0.007*** (0.002) | −0.006** (0.003) | −0.004 (0.003) |
| CEO | 0.113 (0.074) | 0.106 (0.096) | −0.031 (0.143) | −0.089 (0.166) |
| CSR | 0.939*** (0.090) | 0.668*** (0.083) | 0.931*** (0.137) | 1.041*** (0.125) |
| EMT | 0.214** (0.086) | 0.295*** (0.085) | 1.063*** (0.134) | 0.989*** (0.147) |
| INDEP | 0.005*** (0.002) | 0.008*** (0.003) | −0.003 (0.003) | 0.001 (0.004) |
| SUST_INC | 0.351*** (0.085) | 0.280*** (0.070) | 0.158 (0.161) | 0.110 (0.132) |
| Controls | Yes | Yes | Yes | Yes |
| Constant | −2.080*** (0.407) | −2.299*** (0.459) | 1.073* (0.638) | 0.634 (0.616) |
| Observations | 977 | 977 | 458 | 458 |
| R-squared | 0.592 | 0.359 | 0.462 | 0.392 |
| Pooled OLS | Yes | No | Yes | No |
| Random effects | No | Yes | No | Yes |
| Variables | (1) | (2) | (1) | (2) |
|---|---|---|---|---|
| 0.015 | 0.020 | −0.001 (0.006) | 0.005 (0.005) | |
| 0.024 | 0.020 | −0.008 (0.006) | −0.005 (0.005) | |
| −0.007 (0.012) | 0.029 | −0.002 (0.017) | −0.025 (0.021) | |
| −0.009 | −0.007 | −0.006 | −0.004 (0.003) | |
| 0.113 (0.074) | 0.106 (0.096) | −0.031 (0.143) | −0.089 (0.166) | |
| 0.939 | 0.668 | 0.931 | 1.041 | |
| 0.214 | 0.295 | 1.063 | 0.989 | |
| 0.005 | 0.008 | −0.003 (0.003) | 0.001 (0.004) | |
| SUST_INC | 0.351 | 0.280 | 0.158 (0.161) | 0.110 (0.132) |
| Controls | Yes | Yes | Yes | Yes |
| Constant | −2.080 | −2.299 | 1.073 | 0.634 (0.616) |
| Observations | 977 | 977 | 458 | 458 |
| R-squared | 0.592 | 0.359 | 0.462 | 0.392 |
| Pooled | Yes | No | Yes | No |
| Random effects | No | Yes | No | Yes |
This table provides the estimates of the pooled OLS and random-effect regressions for 2011–2022. The global sample is split into two subsamples of listed insurance companies from developed countries (154 companies) and developing countries (98 companies). The dependent variable of interest is the eco-innovation Index. Variables definitions are provided in Table 3. Robust standard errors (SE) are reported in parentheses. ***p < 0.01; **p < 0.05 and *p < 0.10
Third, since the sensitivity and the awareness of specific countries and geographical areas toward sustainability may be different, thus affecting the pursuing of eco-innovation by companies, we control for the environmental performance index (EPI) (Block et al., 2024). The EPI presents a comprehensive data-based overview of global sustainability efforts. By evaluating 180 nations using 58 indicators grouped into 11 key environmental areas, the index assesses progress in climate action, environmental health and ecosystem preservation [1]. These metrics measure how closely each country aligns with predefined environmental policy goals. Serving as both a benchmark and a diagnostic tool, the EPI identifies top performers and those falling behind, offering valuable insights for countries aiming to enhance their sustainability strategies. The findings, reported in Table 9, indicate a positive and statistically significant relationship between national environmental performance (EPI) and firm-level eco-innovation across both model specifications suggesting that firms located in countries with stronger environmental performance tend to achieve higher levels of eco-innovation. The results support the view that environmental policy environments at the national level play a facilitating role in driving eco-innovative behavior, even when controlling for structural differences between countries. With regard to the other variables, the main results are confirmed and remain consistent across both model specifications.
Regressions on eco-innovation index considering the environmental performance index
| Variables | (1) | (2) |
|---|---|---|
| EPI | 0.012*** (0.003) | 0.004* (0.002) |
| GEND | 0.009*** (0.003) | 0.009*** (0.003) |
| BMEET | 0.021*** (0.004) | 0.011*** (0.003) |
| BSIZE | 0.008 (0.012) | 0.008 (0.015) |
| SKILLS | −0.008*** (0.002) | −0.006*** (0.002) |
| CEO | 0.108 (0.079) | 0.112 (0.095) |
| CSR | 1.061*** (0.089) | 0.825*** (0.080) |
| EMT | 0.577*** (0.084) | 0.407*** (0.083) |
| INDEP | 0.004* (0.002) | 0.008*** (0.002) |
| SUST_INC | 0.393*** (0.092) | 0.347*** (0.070) |
| Controls | Yes | Yes |
| Constant | −2.601*** (0.415) | −1.047 (0.451) |
| Observations | 1.092 | 1.092 |
| R-squared | 0.536 | 0.355 |
| Pooled OLS | Yes | No |
| Random effects | No | Yes |
| Variables | (1) | (2) |
|---|---|---|
| 0.012 | 0.004 | |
| 0.009 | 0.009 | |
| 0.021 | 0.011 | |
| 0.008 (0.012) | 0.008 (0.015) | |
| −0.008 | −0.006 | |
| 0.108 (0.079) | 0.112 (0.095) | |
| 1.061 | 0.825 | |
| 0.577 | 0.407 | |
| 0.004 | 0.008 | |
| SUST_INC | 0.393 | 0.347 |
| Controls | Yes | Yes |
| Constant | −2.601 | −1.047 (0.451) |
| Observations | 1.092 | 1.092 |
| R-squared | 0.536 | 0.355 |
| Pooled | Yes | No |
| Random effects | No | Yes |
This table provides the estimates of the pooled OLS and random-effect regressions for 2011–2022. The dependent variable is the eco-innovation index (ECOINN), while the independent variable of interest is the environmental performance indicator (EPI). Variables definitions are provided in Table 3. Robust standard errors (SE) are reported in parentheses. ***p < 0.01; **p < 0.05 and *p < 0.10
4.4 Additional analyses
To enrich our results and their interpretation, we also develop an additional analysis: the survival analysis. Survival analysis (Clark et al., 2003) is a branch of statistics used to predict the time until an event of interest occurs. Key components of survival analysis include the survival function, which represents the probability that the event has not occurred by a certain time, and the hazard function, which describes the instantaneous rate at which the event occurs, given that it has not yet happened. Commonly used methods in survival analysis include the Kaplan−Meier estimator for estimating the survival function, the Cox proportional hazards model for assessing the impact of covariates on the hazard rate and parametric models that assume a specific distribution for survival times. Specifically, we used the Kaplan−Meier estimator (Etican et al., 2017) to estimate the survival function. In medical research, this estimator is often used to measure the fraction of patients who survive for a certain period after treatment. In our analysis, we used this method to categorize insurance companies into groups based on specific criteria: those with an environment management team, those with a CSR sustainability committee and those with sustainability compensation incentives. The dependent variable is a binary indicator, set to 1 if the insurance company achieves and maintains an eco-innovation score higher than 2. This approach enables us to assess how the probability of remaining non-eco-innovative varies over time across different groups of insurance companies.
Upon examining Figure 2 (Panel A), we observe that the probability of remaining non-eco-innovative decreases over time at a slower rate for insurance companies lacking an environment management team. Similarly, as illustrated in Figure 2 (Panel B), this pattern is also evident for insurance companies without a CSR sustainability committee. Furthermore, observing Figure 2 (Panel C), the probability of remaining non-eco-innovative decreases in the short term for insurance companies that have sustainability compensation incentives, although these incentives do not significantly impact the probability in the long term compared to companies without such incentives.
The figure consists of three Kaplan-Meier survival plots. Plot (a) compares companies with and without environmental management teams, showing higher survival probability for those with such teams. Plot (b) analyses the role of a corporate social responsibility committee, with firms having one demonstrating longer survival periods. Plot (c) assesses sustainability compensation incentives, indicating that companies offering such incentives experience slightly improved survival compared to those without. Each plot shows probability of survival on the y-axis against time on the x-axis.Probability of non-eco innovation over time, by presence of environment management team (Panel A), presence of CSR sustainability committee (Panel B) and by presence of sustainability compensation incentives (Panel C)
Note(s): This figure shows the Kaplan−Meier survival estimates of the probability of remaining non-eco-innovative over time. Panel A compares firms with and without an environmental management team (EMT), Panel B compares firms with and without a CSR sustainability committee and Panel C compares firms with and without sustainability compensation incentives. A lower survival curve indicates a faster transition toward achieving an eco-innovation Index above 2, meaning that firms with these governance features tend to adopt eco-innovation practices more quickly than those without
Source: Authors’ own creation
The figure consists of three Kaplan-Meier survival plots. Plot (a) compares companies with and without environmental management teams, showing higher survival probability for those with such teams. Plot (b) analyses the role of a corporate social responsibility committee, with firms having one demonstrating longer survival periods. Plot (c) assesses sustainability compensation incentives, indicating that companies offering such incentives experience slightly improved survival compared to those without. Each plot shows probability of survival on the y-axis against time on the x-axis.Probability of non-eco innovation over time, by presence of environment management team (Panel A), presence of CSR sustainability committee (Panel B) and by presence of sustainability compensation incentives (Panel C)
Note(s): This figure shows the Kaplan−Meier survival estimates of the probability of remaining non-eco-innovative over time. Panel A compares firms with and without an environmental management team (EMT), Panel B compares firms with and without a CSR sustainability committee and Panel C compares firms with and without sustainability compensation incentives. A lower survival curve indicates a faster transition toward achieving an eco-innovation Index above 2, meaning that firms with these governance features tend to adopt eco-innovation practices more quickly than those without
Source: Authors’ own creation
These findings underscore the differential impact of various sustainability governance structures on the eco-innovation trajectories of insurance companies, highlighting the importance of specific organizational commitments to sustainability in driving long-term eco-innovation.
5. Conclusions
This paper investigates the relationship between corporate governance and eco-innovation within the insurance sector. By analyzing a global sample of 252 insurance companies over the period 2011–2022, the study aims to assess how various governance characteristics influence firms’ eco-innovation efforts and outcomes.
The results indicate that governance structures significantly influence how insurance companies perform in terms of eco-innovation. Notably, board gender diversity is positively associated with eco-innovation, indicating that diverse boards contribute to more innovative and sustainable practices. Furthermore, regular board meetings are essential for keeping sustainability at the forefront of company strategies, facilitating agile responses to emerging environmental challenges.
However, the presence of financial expertise on boards appears to hinder eco-innovation, as financially oriented directors tend to prioritize short-term profitability over long-term sustainability goals. This highlights the need for a balanced governance structure that values both financial health and environmental responsibility.
In addition, the presence of dedicated sustainability committees and environmental management teams significantly enhances eco-innovation. These structures ensure that environmental objectives are integrated into the company’s core operations, driving sustained efforts toward sustainability. Similarly, linking executive compensation to sustainability performance further incentivizes eco-innovation, aligning management’s incentives with long-term environmental goals.
From a broader perspective, the study suggests that eco-innovation in the insurance sector is influenced not only by internal governance but also by external factors such as national environmental policies. Firms operating in countries with stronger environmental performance tend to achieve higher levels of eco-innovation, underscoring the importance of supportive institutional frameworks.
Our paper has important implications for both academics and practitioners.
From an academic standpoint, this research enriches the literature on corporate governance and eco-innovation, particularly within the underexplored context of the insurance industry. It extends the eco-innovation index framework developed by Albitar et al. (2023) to a new sector and offers novel insights into the impact of governance on environmental innovation.
For practitioners, especially in the insurance sector, the results underline the importance of governance structures that prioritize sustainability. Managers, investors and regulators can benefit from these findings by recognizing the value of diversity, dedicated sustainability teams and sustainability-linked incentives in fostering eco-innovation. By incorporating these governance practices, insurance firms can not only mitigate environmental risks but also enhance their competitive advantage in a rapidly transitioning economy.
Moreover, from a policy perspective, our findings suggest that regulatory frameworks could benefit from encouraging or mandating governance structures to promote eco-innovation in the insurance sector. For instance, as previously reported, introducing guidelines that contribute to gender diversity on boards or link executive compensation to sustainability targets could accelerate the integration of environmental considerations into corporate strategy. Additionally, policymakers can facilitate the alignment of environmental regulations and incentives with corporate governance reforms to drive sustainable transformation across the insurance industry.
In conclusion, the paper emphasizes that a holistic approach to governance, which balances financial performance with environmental sustainability, is key to achieving meaningful progress in eco-innovation. Insurance companies that embrace such practices are better positioned to contribute to the global transition to a low-carbon economy, ensuring long-term success for all stakeholders involved.
Note
Since the Environmental Performance Index is published on a biennial basis, it does not provide annual data by default. To address this limitation and ensure temporal consistency, we construct yearly values of the EPI for each country by applying an imputation method. This approach allows us to approximate the annual evolution of environmental performance while preserving the underlying trends captured by the original biennial data.

