Does ownership type affect environmental disclosure?


Purpose
In recent years, firms tend to direct their attention in communicating their environmental actions with their stakeholders. However, the level of environmental disclosers varies significantly among firms. This paper aims to explain the variation in environmental disclosure of firms based on their ownership type, namely – state ownership and institutional ownership. The study further aims to understand whether and how the relationship between ownership structure and environmental disclosure changes regarding countries’ development levels.


Design/methodology/approach
This paper uses a sample of 27,847 firm-year observations from 72 countries/economic districts between the years 2002 and 2017 and regression analysis to test how the relationship between different ownership structures and environmental disclosure and whether this relation is conditional on countries’ development levels.


Findings
This study finds that firms with higher state ownership have higher environmental disclosures and higher institutional ownership has a negative effect on environmental disclosures. Furthermore, this paper also documents that firms with higher state ownership and operating in developed countries have incrementally higher environmental disclosure, relative to firms operating in developing countries.


Research limitations/implications
The study has limitations that would provide possible starting points for further research. The first limitation is related to the environmental disclosure measure, which reflects the level of environmental disclosure of firms based on their disclosure information given in the Thomson Reuters, Asset4 database. A more refined measure can be constructed using hand-collected data based on linguistic analysis, which may reflect not only the level of the disclosure but also the quality of the environmental disclosure. The second limitation is the limited focus of the study toward state and institutional shareholding. Therefore, future research may consider examining the different types of ownership such as family ownership.


Practical implications
The findings of the study may help policymakers and regulators to consider the potential impact of various ownership types on environmental disclosures. Also, given the impact of countries’ development levels, regulators should consider that a one-size-fits-all is not applicable in environmental disclosures. Therefore, each country should consider the institutional dynamics of their operating environment to set appropriate regulations to enhance environmental disclosures.


Social implications
From a social perspective, the findings indicate that firms’ stakeholder engagement via environmental disclosures depends on the type of the controlling shareholders.


Originality/value
This study contributes to the literature by developing a new construct for environmental disclosure based on Biodiversity, Climate Change, Environmental Investments and Spill Impact Reduction performance measures. Further, grounding on legitimacy and stakeholder theories, this study shows the influence of ownership type on environmental disclosures and how this effect changes in accordance with the countries’ development.



Introduction
In the past two decades, corporations have been exposed to extensive pressure from society and regulators (e.g. Carbon Disclosure Project (CDP), the Kyoto Protocol of the United Nations Framework Convention on Climate Change) for higher accountability on climate change and environmental issues. As a consequence of these worldwide calls, firms' sensitivity and awareness of environmental issues have increased significantly. Given that firms' reputation and existence in the market at stake (Dintimala and Amril, 2018), in addition to increasing sensitivity and awareness, to provide a positive signal on the market, firms also direct their attention in communicating their environmental actions with their stakeholders. Consequently, more and more firms start with enhanced disclosures on environmental issues. However, firms' environmental disclosures show significant variations. While, previous studies document that environmental disclosure is associated with numerous factors, including the concerns of stakeholders (Ali et al., 2017), firm's strategy and vision, gender diversity (Baalouch et al., 2019), environmental performance (Baalouch et al., 2019;Giannarakis, 2018), size, age, listing status, profitability (Kılıç and Kuzey, 2019a), board independence and existence of a sustainability committee (Kılıç and Kuzey, 2019b) and macro-level dynamics such as political, labor and cultural systems (Baldini et al., 2018), the literature is yet to explain the influence of "ownership" on environmental disclosures. The ownership structure is one of the essential elements of governance mechanisms distinguishing firms' behaviors from one another (Fama and Jensen, 1983) and structured by different determinants. So far, none of the previous studies exclusively examine how ownership types affect environmental disclosure. Thus, this study aims to fill the gap in the literature by analyzing the relationship between different ownership structures and firms' climate change, biodiversity, environmental investments, and spill impact reduction disclosures. This is important because it shows whether institutional and state ownerships could act as a stimulating driver for firms' disclosure policy.
Firms are influenced by a broader social context such as public and private regulations. It is argued that countries that have traditionally been intensely engaged in environmental protection would have more corporate environmental attention, which leads to more disclosure practices (Halme and Huse, 1997). Firms' disclosure practices are affected by country-specific institutional factors such as reporting requirements, political costs of disclosure, cultural and social norms (Meek et al., 1995). Consequently, in such countries, the influence of ownership on environmental disclosure will be affected. Although countries' development level affects voluntary disclosure practices, except few cross-country studies in the area of environmental disclosure (e.g. Calza et al., 2016;Giannarakis et al., 2018) focusing on European firms, the majority of previous studies examining the determinants of environmental disclosures has been studied in a single-country context. For instance, Al Amosh and Mansor (2020) examine the ownership structure on the environmental disclosure level in Jordan, Halkos, and Skouloudis (2016) study the disclosure practices for Greek firms, Iatridis (2013) investigate the relationship between environmental disclosure and corporate governance in Malaysia an emerging market. Given that there is still a lack of cross-country studies examining the changes in the environmental disclosure based on countries' development, we further aim to understand whether and how the relationship between ownership structure and environmental disclosures changes regarding countries' development levels.
Using a sample of 27,847 firm-year observations from 72 countries/economic districts between the years 2002-2017 and regression analysis, we find that: firms with higher state ownership have higher environmental disclosures; higher institutional ownership has a negative effect on environmental disclosures; firms with higher state ownership and operating in developed countries have incrementally higher environmental disclosure, relative to firms operating in developing countries; and firms with higher institutional ownership have a similar level of low environmental disclosure both in developed and developing countries.
Our study has several contributions. First, grounding on legitimacy and stakeholder theories, our findings contribute to the academic literature on environmental disclosure by documenting that ownership type is a significant determinant of firms' environmental disclosure. Furthermore, we show a more comprehensive picture of how the influence of ownership type on environmental disclosures changes in accordance with the countries' development. While prior literature has paid more attention to environmental disclosure in developed and developing countries separately (Ali and Rizwan, 2013;Giannarakis et al., 2018), our study documents the impact of ownership structure on environmental disclosure by focusing on both developed and developing countries together. From a social perspective, our findings indicate that firms' stakeholder engagement via environmental disclosures depends on the type of controlling shareholders. Second, using factor loadings of four environmental disclosure measures climate change, biodiversity, environmental investments, and spill impact reduction, this study contributes to the literature by developing a new construct for environmental disclosure. Third, the findings of our study may help policymakers and regulators to consider the potential impact of various ownership types on environmental disclosures. Also, given the impact of countries' development levels, regulators should consider that a one-size-fits-all is not applicable in environmental disclosures. Therefore, each country should consider the institutional dynamics of their operating environment to set appropriate regulations to enhance environmental disclosures. The remainder of the paper is organized as follows. The following section provides a literature review of environmental disclosure and ownership structure, and it develops hypotheses. The next section introduces the data and methodology used. The fourth section presents the results together with the discussion of the implications of the results. The final section concludes the study, summarizing the main findings and limitations.
Theoretical framework and hypotheses development Environmental disclosure Voluntary disclosure of environmental information aims to minimize information risks and, thereby, their related costs to satisfy stakeholders. Legitimacy theory suggests that firms seek to ensure that their activities and operations are acceptable within the norms of their societies. It posits that the firms must appear to consider the rights of all stakeholders, and thereby, disclosures on corporate social responsibility issues are responses to the pressures from the political, social, and economic environment. Therefore, corporate management may use the annual reports as a corporate response and evidence of sensitivity to specific environmental issues as the awareness and concern in the public increases about related issues (Deegan and Rankin, 1996). The only way to achieve or maintain legitimacy is to inform society about the actions taken, of course, via disclosure (Cormier et al., 2004). Also, while the motivations to disclose environmental information for poor performers are the threat to legitimacy and change the public image, the motivation to disclose environmental information for superior performers is to differentiate themselves from the others (Clarkson et al., 2008).
Similarly, as it is indicated in the stakeholder theory, firms are part of a social system, and stakeholders in that social system have the power to impact their performance. Hence, firms must take action to satisfy the needs of a wide range of stakeholders. Disclosure practices are regarded as one of the critical factors in fulfilling the responsibility to stakeholders (An et al., 2011). Therefore a macro view to disclosure practices is provided by legitimacy theory, whereas a micro view is provided by stakeholder theory, and they both attempt to emphasize that firms must communicate to the whole society to achieve and protect legitimacy and groups in that society from meeting their goals (Cormier et al., 2004).

Environmental disclosure and state ownership
The government, which is one of the influential stakeholders, can influence corporate strategy and performance, thus disclosure practices (Lu and Abeysekera, 2014). It is argued that for the firms when the government holds the majority of the shares, the management of those firms would be willing to disclose more to reflect its social and environmental responsibilities and impact the social perception of the state (Naser et al., 2006;Lan et al., 2013). Other than disclosure of environmentally sensitive activities, the findings of Calza et al. (2016) reveal that firms with state ownership present superior environmental proactivity. Supporting such a view, the results from numerous previous studies have shown a positive relationship between the level of voluntary disclosure and state ownership (Eng and Mak, 2003;Naser et al., 2006;Wang et al., 2008;Haddad et al., 2015;Khlif et al., 2017). Amran et al. (2012) also found a significant positive relation between climate change disclosure practices and firm size, profitability, government ownership, and business network. Due to the voluntary nature of environmental disclosure and in line with legitimacy and stakeholder theories, we argue that state ownership increases such disclosure practices, as governmental members on the board are more likely to demonstrate their efforts about environmental sensitivity. Hence, we propose the following hypothesis:

H1.
Voluntary environmental disclosure on climate change, biodiversity, environmental investments, and spill impact reduction is higher for firms with higher state ownership.
Environmental disclosure and institutional ownership Institutional ownership refers to the situation where the largest shareholder is an institution or not. As institutional owners are more sophisticated and experienced with access to relevant information (Balsam et al., 2002), it is suggested that they would be more effective in controlling and monitoring management's activities (Siregar and Utama, 2008). Generally, large institutional investors enjoy the privilege to access internal sources of information, not available to all shareholders (El-Diftar et al., 2017). Thus, such a privilege causes them to avoid more voluntary disclosure about social and environmental issues. Empirical evidence indicates a negative relation between institutional ownership and voluntary disclosure (Tsamenyi et al., 2007;Samaha et al., 2012;Juhmani, 2013). For example, Siew et al. (2016) investigated the effect of environmental, social, and governance disclosures and institutional ownership on information asymmetry. Despite their results showed environmental, social, and governance disclosures and institutional ownership reduce information asymmetry, further analysis indicated that higher levels of institutional ownership weaken this negative relationship. In a similar vein, the findings of Bushee et al. (2004) reveals firms are less likely to make disclosures when institutional owners dominate them. One potential reason for such a negative impact of institutional owners on environmental disclosure can be explained via their shifted focus toward firms' financial short-terms interest rather than long-term sustainable growth. Given institutional owners focus more on short-term financial benefits, they are less sensitive to the needs of the society and other stakeholders and less likely to legitimize their actions via voluntary disclosures to send a positive signal to the market. Therefore, in line with the legitimacy and stakeholder theories, we argue that there is a negative relation between institutional ownership and environmental disclosure. Hence, we propose the following hypothesis:

H2.
Voluntary environmental disclosure on climate change, biodiversity, environmental investments, and spill impact reduction is lower for firms with higher institutional ownership.
The moderating role of countries' development Countries have different institutional settings. The differences in countries' institutional settings consequently affect firms' organizational structure and forms, including the shareholders. In terms of environmental disclosure and ownership relation, the impact of institutional setting can be explained by grounding on the stakeholder and legitimacy theory. For example, in countries where the stock markets are developed and stakeholders have a more active role, shareholders are likely to direct their attention to the needs of the society and tend to send a positive signal to the market to ensure the information need of all stakeholders. On the other hand, in countries with less developed stock markets and institutional settings, the shareholders will be less sensitive to the needs of society. In line with stakeholder and legitimacy theories, we argue that the impact of shareholders, therefore, depends on the level of institutional development in a country. With this respect, country specific-factors such as development status and disclosure practices have received attention from many researchers in recent years (Amran et al., 2014). It is argued that countries that have traditionally been intensely engaged in environmental protection would have more corporate environmental attention, which leads to more disclosure practices (Halme and Huse, 1997). There are different incentives for developed and developing countries regarding their environmental disclosure. For instance, while Ali et al. (2017) found that developed countries' corporate social responsibility disclosure practices are shaped by specific stakeholders such as regulators, shareholders, creditors, investors, environmentalists, and the media; they found that developing countries' influenced by external forces such as international buyers, foreign investors, international media and international regulatory bodies. Baldini et al. (2018) studies country-level determinants together with firm-level determinants and revealed that political, labor and cultural systems affect firms' environmental, social, and governmental disclosure practices. In addition, all industrialized firms except the USA ratify the Kyoto Protocol, which sets a limit on the amount of greenhouse gas emissions. Therefore, most of the firms in developed countries receive pressure from the government to follow carbon emission regulations and tend to disclose more about climate change. We, therefore, argue that when compared to developing ones, are more environmentally sensors and would be more engaged in environmental disclosure practices. Thus, we propose the following hypotheses: H3. Firms with high state ownership have incrementally higher environmental disclosure on climate change, biodiversity, environmental investments, and spill impact reduction in developed countries relative to developing countries.
H4. Firms with high institutional ownership have incrementally higher environmental disclosure on climate change, biodiversity, environmental investments, and spill impact reduction in developed countries relative to developing countries.

Data
In this study, we draw our sample from the full international universe of the Thomson Reuters Asset4 database with available environmental data about Biodiversity Impact Reduction, Climate Change Risks and Opportunities, Environmental Investments Initiatives, and Spill Impact Reduction over the period 2002-2017. The initial sample has a population of 55,442 firm-years. Furthermore, we collected data for state and institutional ownership and all other financial and non-financial (ownership and corporate governance) data from Datastream, Worldscope, and Thomson Reuters Asset4 databases. After eliminating firms with missing data for any of the independent and control variables, we have a final sample of 27,847 firm-year observations.

Research model
To examine H1, voluntary environmental disclosure on climate change, biodiversity, environmental investments, and spill impact reduction is higher for firms with higher state ownership, and H2, voluntary environmental disclosure on climate change, biodiversity, environmental risk, and spill impact reduction is lower for firms with higher institutional ownership, we use the equation (1):

þIndustry; Auditor; Year and Country indicators þ «
Where ln(EnvDisclosure) is the natural logarithm of environmental disclosure, a higher value of environmental score indicates better environmental disclosure. It is computed using factor loadings of four environmental disclosure measures BioDiversity, ClimateChange, Env_Investments, and SpillImpactReduction. To avoid inter-correlated components within the newly created disclosure quality index variables, we rotated the factors using the orthogonal varimax method (Kaiser, 1958). Table 1 presents the details of the factor loadings. Biodiversity is an indicator variable that equals one if the firm reports on its impact on biodiversity or on activities to reduce its impact on the native ecosystems and species, as well as the biodiversity of protected and sensitive areas. ClimateChange is an indicator variable that equals one if the firm reports about its awareness of commercial risks and opportunities of climate change. Env_Investments is an indicator variable that equals one if the firm reports on making proactive environmental investments or expenditures to reduce future risks or increase future opportunities. Finally, SpillImpactReduction is an indicator variable that equals one if the firm reports on initiatives to reduce, avoid or minimize the effects of spills (environmental crisis management system or disaster recovery plan).
We expect that ln(StateOwn(%)) will have a positive and significant coefficient, indicating that firms with higher state ownership have higher environmental disclosure. On the other hand, we expect the coefficient of ln(InstitutionalOwn(%)) to be a negative and significant coefficient, suggesting that environmental disclosure is lower for firms with higher institutional ownership. Furthermore, to test the prediction of H3 and H4, the moderating role of countries' development on the association between voluntary environmental disclosure and state and institutional ownership, respectively, we use equation (2). In equation (2), our main variable of interest is the interaction term ln(Own(%)) X Developed. Similar to equation (1), in equation (2), ln(Own(%)), takes two different variables to test H3 and H4, respectively; (1) ln(StateOwn(%)) and (2) ln(InstitutionalOwn(%)).
Developed is an indicator variable that equals one if the country where the firm is originated is a developed country, 0 otherwise. We expect the coefficients of ln(StateOwn(%)) X Developed to be positive and significant, indicating that the positive association between environmental disclosure and state ownership is incrementally higher in developed relative to developing countries. Similarly, we expect a positive and significant coefficient for ln(InstitutionalOwn(%)) X Developed, suggesting that the negative association between environmental disclosure and institutional ownership is weakening in developed relative to developing countries. Moreover, we use firm-specific financial and non-financial (ownership and corporate governance) controls in both equations. Firms' strategic and operational decisions are highly influenced by their financial strength. Therefore, to control for firms' financial performance and situation, we use the size of the firms (FirmSize), leverage (DebttoCapital), Return on Equity (ROE), and cash flows from operations (CFO) as financial controls. Likewise, firms' decisions over environmental disclosures are more likely to be determined by the management, more specifically the CEO and the monitoring bodies, mainly the board of directors and corporate governance. Thus, to control the impact of CEO and firm-level corporate governance on environmental disclosures, we use Reputation_Monitoring, CEO_BoardMem, CEO_Comp_Link, CG_Comm CSR_Comm, and CloselyHeldShares(%). Table 2 defines all variables used in our analyzes.
Finally, we control for the potential impact of auditors, industry, year, and country on the environmental disclosures by including auditors, industry, year, and country indicators. In all our estimations, we use Huber/White/sandwich standard error estimates clustered by firms to correct potential heteroskedasticity and within-cluster correlation. Table 3 presents descriptive statistics, which are also separately presented by developed and developing countries. Additionally, univariate test statistics for the mean differences between developed and developing countries are presented. In our sample, Ln (EnvDisclosure) has a mean value of À0.190, with a minimum value of À0.668 and a maximum value of 1.297. Developed countries are having a lower score (À0.196), and developing countries having a higher score (À0.158). Among the four dimensions of environmental disclosure, the highest belongs to climate change disclosure that is 32.7% for the whole sample. Also, developed countries have a greater climate change disclosure score, which is 33.2%, and developing countries have a lower one, which is 30%.

Descriptive statistics and univariate analysis
Moreover, the mean values of state ownership for developed and developing countries are 0.105 and 0.604, respectively. This result reveals that developing countries have greater state ownership. Also, the mean values of institutional ownership for developed and developing countries are 1.278 and 0.546, respectively. This result reveals that developed countries have greater institutional ownership. Table 4 presents the Pearson correlations. The correlation coefficients assure that multicollinearity is not a severe problem for the variables, as the correlation coefficients do Table 2.

Ln(EnvDisclosure)
The log of the weighted index computed using factor loadings of four environmental disclosure measures BioDiversity, ClimateChange, Env_Investments and SpillImpactReduction (please see Table 1 for further details BioDiversity_indicator An indicator variable that equals 1 if the firm reports on its impact on biodiversity or on activities to reduce its impact on the native ecosystems and species, as well as the biodiversity of protected and sensitive areas, 0 otherwise (Asset4 ENERDP019) ClimateChange_indicator An indicator variable that equals 1 if the firm reports about its awareness on commercial risks and opportunities of climate change, 0 otherwise (Asset4 ENERDP089) Env_Investments_indicator An indicator variable that equals 1 if the firm reports on making proactive environmental investments or expenditures to reduce future risks or increase future opportunities, 0 otherwise (Asset4 ENERDP095) SpillImpactReduction_indicator An indicator variable that equals 1 if the firm reports on initiatives to reduce, avoid or minimize the effects of spills (environmental crisis management system or disaster recovery plan), 0 otherwise (Asset4 ENERDP087)

Ln(StateOwn(%))
The log of the percentage of shares held by the government (Datastream NOSHGV)

Ln(InstitutionalOwn(%))
The log of the percentage of shares held by investment firms (Datastream NOSHIC) Developed An indicator variable that equals 1 if the country where the firm originated is a developed country, 0 otherwise (UNCTAD, United Nations Conference on Trade and Development)

FirmSize (lnTotalAssets)
The log of the total assets (Worldscope WC02999 in US$)

Leverage (lnDebttoEquity)
The log of the total debt to total equity ratio (Worldscope WC08231)

ROE (lnROE)
The log of the return on equity (Worldscope WC08301)

CFO (lnCFO)
The log of the cash flow from operations (Worldscope WC04860) Reputation_Monitoring_indicator An indicator variable that equals 1 if the firm monitors its reputation or its relations with communities, 0 otherwise (Asset4 SOCODP021) CEO_Comp_Link_indicator An indicator variable that equals 1 if the CEO's compensation is linked to total shareholder return, 0 otherwise (Asset4 CGCPDP041) CEO_BoardMem_indicator An indicator variable that equals 1 if the CEO is a member of the board of directors, 0 otherwise (Asset4 CGBSDP061) CG_Comm_indicator An indicator variable that equals 1 if the firm has a corporate governance committee, 0 otherwise (Asset4 CGBFDP005) CSR_Comm_indicator An indicator variable that equals 1 if the firm has a corporate social responsibility committee, 0 otherwise (Asset4 CGVSDP005) Ln(CloselyHeldShares(%)) The log of the percentage of shares held by insiders (Worldscope WC08021)  Notes: This table presents the summary statistics for the full sample, developed and developing countries sample. The mean difference test is conducted between the developed and developing firms for each variable. All variables are described in Table 2. T-values are presented in parentheses. ***, ** and * denote the significance level at 1%, 5% and 10% Table 3. Ln(StateOwn(%)) 6 0.11 0.13 0.06 0.07 0.05 1.00 Ln(InstitutionalOwn(%)) 7 À0.11 À0.10 À0.05 À0.10 À0.06 À0.13 1.00 FirmSize ( Ln(CloselyHeldShares(%)) 17 À0.07 À0.01 À0.13 0.00 À0.03 0.16 À0.28 À0.09 À0.08 À0.03 À0.09 À0.05 À0.40 À0.15 À0.38 À0.08 Table 4. Correlation matrix not exceed 0.50 for most of the variables. There is a positive association between environmental disclosure and state ownership, especially for biodiversity disclosure; this positive association is more powerful. On the contrary, there is a negative association between environmental disclosure and institutional ownership. Table 5, Model 1, presents the regression results for the relationship between state ownership and environmental disclosure. In line with H1, the coefficient of ln(StateOwn(%)) is positive and significant (b = 0.030; t-values = 2.901), suggesting that firms with higher state ownership have higher environmental disclosure. To test the robustness of our findings and the sensitivity of our results to model specification, we also used a random effect panel regression. Untabulated results indicate that our results are statistically similar to our findings presented in our main analyzes. Furthermore, Table 5, Model 2, presents the impact of institutional ownership on environmental disclosure. The coefficient of ln(InstitutionalOwn(%)) is negative and significant (b = À0.014; t-values = À3.322), indicating that firms with higher institutional ownership have lower environmental disclosure, which is in line with our arguments in H2.

Main findings
Finally, Table 5, Model 3, presents coefficient estimates of both state and institutional ownership in a single model. Our findings in Model 1 and Model 2 remain statistically the same. In summary, our results show that firms' environmental disclosures change according to the ownership type. While firms with higher state ownership have higher environmental disclosures, firms with higher institutional ownership have lower environmental disclosures. Table 4, Model 1, presents the regression results of the incremental effect of state ownership on the environmental disclosure in developed countries. The interaction term, ln (StateOwn(%)) X Developed, captures the incremental effect of developed countries on the Notes: All variables are described in Table 2. T-values are presented in parentheses. ***, ** and * denote the significance level at 1%, 5% and 10% association between state ownership and environmental disclosure. In line with H3, the coefficient of ln(StateOwn(%)) X Developed is positive and significant (b = 0.029; t-values = 4.221). This result indicates that in developed countries, the positive association between state ownership and environmental disclosure is strengthened. To visualize the moderating impact of countries' development level, in Figure 1, we present the margins plot.
As it is seen in Figure 1, firms with higher state ownership and operating in developed countries have environmental disclosure. Similarly, Table 6, Model 2 documents the regression results of the incremental effect of institutional ownership on the environmental disclosure in developed countries. The interaction term, ln(InstitutionalOwn(%)) X Developed, captures the incremental effect of developed countries on the association between institutional ownership and environmental disclosure. Unlike our expectations in H4, the coefficient of ln(InstitutionalOwn(%)) X Developed is insignificant. Therefore, we do not have any support to conclude that the impact of institutional ownership on environmental disclosure is less harmful in developed countries relative to developing countries. We also documented the moderating results in Figure 2, where there is no significant pattern between developed and developing countries in terms of the association between environmental disclosures and institutional ownership.
Further, to increase the robustness of our results, in Table 6, Model 3, we re-estimate Model 1 and Model 2 by incorporating both state and institutional ownership into the same model. Our results stay statistically the same as in Models 1 and 2. Overall, our results suggest that while the impact of state ownership on environmental disclosure is moderated by the development of the countries where the firms are operating, and while country development strengthens the impact of state ownership on environmental disclosure, countries development does not mitigate the negative association between institutional ownership and environmental disclosure.  Notes: All variables are described in Table 2. T-values are presented in parentheses. ***, ** and * denote the significance level at 1%, 5% and 10% Notes: All variables are described in Table 2. T-values are presented in parentheses. ***, ** and * denote the significance level at 1%, 5% and 10% Table 7. Additional analysis IJCCSM Additional tests In Table 7, we further examine the impact of ownership type on environmental performance rather than environmental disclosures like in our primary analysis. Using a smaller sample where data is available, we replaced our dependent variable disclosure with: CO 2 equivalents emission total; Environmental provisions; Emission reduction/waste recycling ratio; and Waste recycling ratio.
Our results show that while state ownership has a significant impact on environmental performance, institutional ownership does not have a significant influence.

Conclusion
In this study, we have examined the influence of firms' ownership structures on their level of environmental disclosure. Using the assumptions of legitimacy theory and stakeholder theory, we have tried to extend the previous literature on the relationship between voluntary environmental disclosure level and ownership structure considering institutional ownership and state ownership. Moreover, we contribute to previous literature on voluntary environmental disclosure and country-specific factors considering their development status. Our main findings suggest a positive relationship between state ownership and voluntary environmental disclosure and a significant negative relationship between institutional ownership and voluntary environmental disclosure. So, the firms with higher state ownership have higher environmental disclosure scores, and the firms with higher institutional owners have lower disclosure scores for climate change, biodiversity, environmental risk, and spill impact reduction.
Furthermore, the findings suggest that the positive relationship between state ownership and environmental disclosure is higher in developed countries. However, we find no evidence about the effect of institutional ownership on environmental disclosure is less harmful in developed countries relative to developing countries.
Our study has some limitations that would provide possible starting points for further research. The first limitation is related to our environmental disclosure measure, which reflects the level of environmental disclosure of firms based on their disclosure information given in the Thomson Reuters, Asset4 database. A more refined measure can be constructed using hand-collected data based on linguistic analysis, which may reflect not only the level of the disclosure but also the quality of the environmental disclosure. Using hand-collected data and contracting it for a comprehensive and international sample is not feasible due to practical reasons (e.g. language barriers to conduct the linguistic analysis in different countries, limited access to local data, judgmental differences among researchers). The second limitation is the limited focus of our study toward state and institutional shareholding. Therefore, future research may consider examining the different types of ownership such as family ownership.
Despite the limitations, our study offers theoretical, practical, and social contributions. On the theoretical side, the study extends prior research in the field of environmental disclosure and ownership type by focusing on both developed and developing countries together, implying that the type of ownership is a significant determinant of the level of firms' environmental disclosure and this effect is conditional on countries development. From a practical perspective, the results of the study would make policymakers and regulators aiming to manage potential ownership types considering their impact on environmental disclosures. Specifically, our results indicate that each country should consider the institutional dynamics of the countries while setting regulations to enhance environmental disclosures. Finally, from a social perspective, our findings support the idea that firms' stakeholder engagement via environmental disclosures depends on the type of controlling shareholders.