Corporate Environmental Disclosure Practices in Different National Contexts: The Influence of Cultural Dimensions

The influence of different national contexts, including the effects of cultural environments, on corporate environmental disclosure practices has yet to be properly addressed in the literature. The purpose of this research is, therefore, to analyse how cultural factors affect the environmental disclosure practices of companies in different countries. This research is supported by the diversity of cultures across countries. Given that a cultural framework prompts different organisational actions and strategies, the question to be answered through this research is as follows: How do cultural aspects affect corporate environmental disclosure? Cultural factors are precisely those that can explain similarities and differences between stakeholders’ actions and preferences. The sample used in this research comprises companies in 28 countries and 9 economic sectors for the period 2004 to 2015. Our main findings show that companies operating in countries with individualist, masculine and indulgent cultures are less likely to disclose environmental information. Contrary to our predictions, cultures with a long-term orientation also discourage the reporting of environmental information, while uncertainty avoidance contexts tend to promote more environmental reporting.


Introduction
Environmental issues are now one of the biggest concerns for companies and society in general. Many companies have been criticised for their negative impact on the environment, and it is therefore logical that companies are interested in showing society that they are concerned about its welfare. To do so, they resort to the disclosure of environmental information, both qualitative and quantitative (Burritt, 2002;Cho & Patten, 2007;Craig & Diga, 1998;Vormedal & Ruud, 2009). At the same time, stakeholders are increasingly putting pressure on companies to introduce major environmental responsibility measures, such as publishing environmental reports and, thereby, disclosing aspects related to their public image, activities and aspirations associated with the environment (Berthelot, Cormier, & Magnan, 2003;Kolk, 2003).
Among the various factors affecting environmental disclosure, the diversity of cultures across countries is emphasised. Differences in national cultures have very important implications for ethics, corporate social responsibility (CSR), organisational culture and managerial practices. Given that a cultural framework affects on the way companies behave (Williamson, 2000) and prompts different organisational actions and strategies (Scott, 2008), the question to be answered through this research is as follows: How do cultural aspects affect corporate environmental disclosure?
In view of the above ideas, our research purpose is to shed more light on the impact that different cultures have on these disclosure practices (Scott, 2001). To do so, we follow the dimensions proposed by Hofstede (1983), Hofstede and Hofstede (2005) and : power distance, individualism, masculinity, uncertainty avoidance, long-term orientation and indulgence. These cultural dimensions have been tested in previous studies in the field of CSR (Orij, 2010;Van der Laan Smith, Adhikari, & Tondkar, 2005) and, to a lesser extent, in the environmental field.
In general, our main findings show that certain cultural dimensions, such as individualism, masculinity and indulgence, have a negative effect on environmental information disclosure. In contrast, power distance does not affect environmental reporting, while uncertainty avoidance and long-term orientation have positive and negative effects, respectively, on the disclosure of environmental issues, contrary to our expectations in both cases.
Our findings also make several contributions to the debate on environmental disclosure. First, the effect that cultural contexts have on environmental disclosure is analysed, building on previous research studies (De Villiers & Marques, 2016;Kolk & Perego, 2010;Once & Almagtome, 2014;Semenova & Hassel, 2016). The study delves into the type of information international companies disclose about their environmental practices by considering 53 items on three related matters: innovation, resource use and emissions. Second, prior literature has drawn on different theoretical approaches and foundations, such as institutional theory and stakeholder theory, to analyse the relationship between cultural dimensions and environmental reporting (e.g., García-Sánchez, Cuadrado-Ballesteros, & Frias-Aceituno, 2016;Orij, 2010;Roy & Goll, 2014). Additionally, most of this past research has used Hofstede's cultural dimensions as a cultural framework for examining such a relationship, showing mainly that the effect of Hofstede's cultural dimensions on environmental reporting is as expected by theory. These cultural dimensions are power distance, uncertainty avoidance, individualism, masculinity, long-term orientation and indulgence. Our research is also based on Hofstede's cultural framework, in line with most of the earlier research on cultural dimensions, but we do not consider institutional and stakeholder theories. Our evidence shows that the power distance dimension does not affect environmental reporting, which contradicts past research that supports a negative association. Furthermore, our findings also demonstrate that uncertainty avoidance and long-term orientation have a positive and negative effect, respectively, on the disclosure of environmental information, which is also contrary to most of the previous research on this issue. Thus, our findings on the power distance, uncertainty avoidance and long-term orientation dimensions contribute to existing theory on cultural and environmental practices in organisations. The discouragement of environmental reporting in cultures with a longterm orientation is the most novel finding of this research because it counters what prior research has stated regarding managers' temporal orientation. Our evidence may reconcile the competing views regarding national cultures and environmental reporting by bringing together preceding contradictory results concerning cultural contexts and environmental disclosure. Our conclusions propose new possibilities of research for Hofstede's cultural framework and environmental disclosure literature. Finally, based on the results obtained, it is evident that the cultural context in which companies operate is influential and that environmental reporting is determined by cultural pressures, as suggested by scholars such as Aerts, Cormier, and Magnan (2006).
The article is structured as follows: First, the theoretical framework is explained. Then, the hypotheses are presented. Next, the methodology and the variables used are outlined. Then, the results are analysed. Finally, the conclusions and our findings' implications are presented.

National Culture and Environmental Disclosure
In recent years, a growing number of organisations have sought to modify their environmental practices so as to reduce their environmental impacts and improve their corporate image (Barnett, Darnall, & Husted, 2015;Monteiro & Aibar-Guzmán, 2010). In addition, companies have also begun disclosing corporate environmental information to avoid potential conflicts with their stakeholders (Epstein & Roy, 2001). Scholars such as Cerin (2002), Berthelot et al. (2003), Moneva and Ortas (2010), and Higgins and Larrinaga (2014) contend that stakeholders require organisations to act in an eco-friendly way; and thus, many companies throughout the world have now increased the level and scope of their environmental information disclosure as a value-added tool (Gamble, Hsu, Kite, & Radtke, 1995). Several reasons may be considered to explain why organisations report on environmental issues. Scholars such as Deegan and Samkin (2006) consider that an initial reason may be that companies wish to show how responsible they are towards a wide range of stakeholders. Through environmental disclosure, companies respond to stakeholders' expectations and, in so doing, contribute to the welfare of society (Hooghiemstra, 2000).
Others argue that companies report on environmental aspects because they are forced to do so by various cultural pressures, as they operate within contexts formed by national cultures that affect their behaviour and impose certain expectations on them (Campbell, 2007;Collier, 2001;Modell, 2002). This contextual relationship means that companies operating in countries with similar cultural dimensions will adopt similar behaviours. In this context, organisations operating in similar environments tend to adopt the same strategic behaviour, and it thus focuses on the deeper aspects of social structures (Claessens & Fan, 2002). According to Scott (2008), a cultural system can be considered an example of a social structure because it introduces a dimension of social life that is prescriptive, evaluative and obligatory. According to Tsakumis (2007), cultural systems explain the similarities and differences of how the human mind is programmed, which in turn differentiates societies around the world. As these cultural aspects differ, so do the ways in which organisations operate (Hofstede, 1980). As indicated by Richardson and Boyd (2005), companies operating in different cultural systems are forced to adopt sustainable behaviour that will mould their standards of transparency and environmental practices.
Several studies have analysed the impact that cultural context has on environmental issues (Adams & Kuasirikun, 2000). This may be because cultural aspects are significant when assessing the importance of environmental issues as, generally speaking, stakeholders in different countries have varying expectations regarding corporate decisions due to their differing cultural conditions that inform different personal values, standards and practices (Carroll, 1979). The research conducted by Hoffman (1999) also focuses on cultural dimensions and specifically builds a framework for understanding the coherence of organisational fields and cultural contexts in environmental matters for companies operating in the U.S. chemical industry.
Considering companies in Canada and the United States, Buhr and Freedman (2001) explore the role cultural factors play in environmental disclosure. They contend that Canadian culture is more propitious for corporate environmental disclosure than U.S. culture because Canadian society has a more collectivist nature that induces companies to report this kind of information. Roy and Goll (2014) examine the influence of national culture on various facets of a country's sustainability indicators-namely, environmental aspects. Environmental aspects refer to the protection of nature. Data from 57 countries were used in the analyses with countries from all over the world, various forms of government and all population sizes. The results obtained support the basic argument that cultural practices influence environmental behaviours.
A culture's impact tends to be analysed through the dimensions proposed by Hofstede (1980Hofstede ( , 2001, Hofstede and Hofstede (2005), and . Although the model initially considered four cultural dimensions, two more were subsequently included, thereby totalling six dimensions. These dimensions provide a major framework not only for analysing a national culture but also for considering the effect cultural differences have on management and organisation. According to Beugelsdijk, Maseland, and van Hoorn (2015), Hofstede's framework predominates in analysis of national cultures in international business and management and represents a relevant source for the investigation of national cultural issues.
This research therefore aims to verify how corporate environmental disclosure responds to different cultural systems. Hofstede's six cultural dimensions are used as proxies for cultural contexts, consistent with past investigations that have also employed them (Caprar, Devinney, Kirkman, & Caligiuri, 2015).

Research Hypotheses
Organisations are influenced by cultural aspects that reflect the way in which organisations conform to the norms, values and cultures that distinguish one society from another throughout the world (Tsakumis, 2007). Hence, culture guides the behaviour of the members of society by providing a dominant logic (Roy & Goll, 2014). Hofstede (2001) considers culture to be a "collective programming of the mind that distinguishes the members of one group or category of people from another" (p. 10).
Several models have been developed to understand cultural differences, such as those by Schwartz, GLOBE and Hofstede. The last one is the most widely used in the literature on different business topics such as accounting, business, management and economics (De Mooij & Hofstede, 2010). In addition, Hofstede's model of cultural dimensions has also been used in various studies of macrocultural contexts related to issues of social responsibility and business ethics (García-Sánchez, Rodríguez-Ariza, & Frías-Aceituno, 2013;Ringov & Zollo, 2007;Williams & Zinkin, 2008), although macrocultural contexts in the environmental field have been less studied.
Currently, there are six cultural dimensions (Hofstede, 1983;Hofstede & Hofstede, 2005): power distance, individualism, masculinity, uncertainty avoidance, longterm orientation and indulgence. These dimensions refer to different aspects ranging from the acceptance of women's role and one's idea of society to the assignment of values orientated to the past, present, or future.
Several research studies have analysed how different cultural systems influence the field of CSR. Van der Laan Smith et al. (2005) conducted a study on U.S. and Scandinavian companies using three of Hofstede's six cultural dimensions. The authors conclude that companies in Scandinavian countries have a greater commitment to social and environmental practices than their U.S. counterparts. Orij (2010) also focused on companies belonging to different countries, analysing the relationship between corporate social disclosure and national culture. Other studies, such as those conducted by Vachon (2010) and Husted (2005), have analysed the impact of national cultures on corporate social disclosure, reporting contradictory findings attributable to different cultural environments.
In the environmental field, Fekrat, Inclan, and Petroni (1996) conducted their research on a sample of 168 companies operating in 18 countries with different cultural systems. Their research reveals differences in the environmental information disclosed by these companies. Gamble, Hsu, and Tollerson (1996) also found significant differences in environmental practices for a sample of 276 companies in 27 countries. Once and Almagtome (2014) analysed Hofstede's cultural dimensions for companies from 20 countries, finding that some of the dimensions, such as individualism and long-term orientation, are related to greater corporate environmental disclosure, while a high degree of power distance is related to less disclosure.
Considering the above, research hypotheses are formulated for each of Hofstede's cultural dimensions to analyse the behaviour of environmental disclosure in different cultural contexts.

Power Distance
One of the cultural dimensions established by Hofstede (2001) describes the level of hierarchy in a society, known as power distance. This dimension represents the extent to which members of organisations within a given culture expect and accept that power is distributed unequally. Inequality might be apparent in the difference between what individuals bring to society and what they receive from it in the distribution of power or in each individual's rights and obligations. According to Waldman, de Luque, Washburn, and House (2006), cultures with greater power distance accept that the hierarchy between superiors and subordinates is extensive and legitimate. Hofstede and Hofstede (2005) consider that when there is a great power distance, positions are placed vertically, giving rise to different levels of power; individuals with less power have less interest in social rights. Accordingly, Miska, Szöcs, and Schiffinger (2018) contend that people in cultures with a great power distance tend to group themselves into different classes depending on a range of criteria. Power bases tend to be stable, and the belief is that power ensures social order, relational harmony and stability. In such cultures, only a few people have access to resources, capabilities and skills. This means the practices of power distance correlate negatively with economic prosperity, competitiveness and human development.
Regarding social and environmental disclosure by organisations, Veser (2004) has found that when power distance is high, stakeholders consider it less likely they will receive a large amount of information on corporate social and environmental practices. However, when there is a lower power distance, organisations need to disclose more information about these practices to gain stakeholder benefits/approval. Similarly, Gray (1988) considers that as power distance increases, information disclosure decreases, because power inequalities are preserved and stakeholders are less likely to have higher expectations of social and environmental disclosure as they comply with their home culture and believe that power should be concentrated in the hands of a few.
Previous studies on power distance and corporate social and environmental disclosure have not obtained unanimous results (Miska et al., 2018). In this respect, Ho, Wang, and Vitell (2012) and Ioannou and Serafeim (2012), for example, have shown that greater power distance increases the disclosure of environmental information, while Orij (2010) and Peng, Dashdeleg, and Chih (2014) have found a negative relationship between power distance and corporate environmental practices. These results are in line with those indicated by Vachon (2010), who considers that companies in countries with less power distance will be less concerned about relations with shareholders and will feel more responsible for the welfare of the community at large and for publicising their environmental practices. To analyse this relationship, the following hypothesis is proposed: Hypothesis 1: The higher the power distance in a society is, the lower or weaker the environmental disclosure will be.

Individualism
Another dimension established by Hofstede (2001) is individualism. In an individualist society, people feel comfortable when empowered to make a decision based on what they think is best, and freedom and independence of the individual is considered of great importance, with priority being given to personal interests over those of social groups. Furthermore, in this type of society, systems protecting an individual's rights are highly developed, and stakeholders are less interested in achieving objectives that are not their own. As a result, companies in a cultural environment with an individualist nature will be less willing to disclose social and environmental information.
In contrast, a collectivist society is formed by individuals who think more as members of a group than as individuals and, therefore, have stronger links to society (Hofstede & Hofstede, 2005). In short, this type of society has close ties between individuals, extended families and groups, whereby everyone accepts responsibility for the members of their group (Peng & Lin, 2009). In this type of society, the group is considered more important than the individual, and the values that predominate are cohesion and consensus, with personal and private initiatives being seen as secondary (Scholtens & Dam, 2007). Greater emphasis is therefore placed on the impact of organisations on society, and collectivist societies tend to be more sensitive to the interests of stakeholders (Blodgett, Lu, Rose, & Vitell, 2001;Ho et al., 2012).
Previous studies, such as those conducted by García-Sánchez et al. (2016), show that organisations belonging to countries where a collectivist culture prevails tend to have greater incentives to disclose social and environmental information to their stakeholders to aid their decision-making processes. Thus, considering the above arguments and the previous evidence, we propose the following hypothesis: Hypothesis 2: The more individualist a society is, the lower or weaker its corporate environmental disclosure will be.

Masculinity
The masculinity dimension refers to gender and the role of women in society. Male-orientated cultures tend to be more assertive and focused on material success, while those with a female orientation tend to be more cooperative, modest and focused on quality of life. Societies that consider themselves masculine describe men as assertive, aggressive, ambitious, competitive and materialistic, while cooperative behaviour is less appreciated.
In some societies, the social roles of men and women overlap, with neither sex behaving competitively, while other societies are considered to have a female orientation. Hofstede (1980) considers that female-orientated cultures reveal a preference for cooperation, constraint, care for the weak and quality of life. In general, organisations with a feminine culture are not as competitive as those with a masculine one, as the former give more priority to concern for others and little distinction is made between men and women in the same position (Hofstede, 2001). According to Orij (2010), masculinity is the opposite of a culture's social orientation, whereby less masculine societies have a greater orientation towards stakeholders (Van der Laan Smith, Adhikari, Tondkar, & Andrews, 2010).
Scholars such as Peng et al. (2014) observe that cultures with a high degree of masculinity place more importance on goals such as a professional career and business success. In contrast, people in cultures where masculinity is not so important give greater value to the group and society (Gray, 1988), while stakeholders seek information about corporate decisions, such as those related to preservation of the environment and community development.
Regarding the results obtained in previous research, there seems to be a negative relationship between masculinity and corporate social and environmental practices (Husted, 2005;Orij, 2010). Other researchers have also found a negative relationship between masculinity and environmental sustainability, noting that the greater the degree of femininity of a given culture, the greater the degree of sustainability, environmental management and commitment to sustainable development (Peng & Lin, 2009;Roy & Goll, 2014). Therefore, based on the previous arguments, we propose the following hypothesis: Hypothesis 3: The more masculine a society is, the lower or weaker its corporate environmental disclosure will be.

Uncertainty Avoidance
This dimension measures how members of a culture feel threatened by uncertain, unknown, or unstructured situations and represents the level of aversion to the unknown. According to Sully de Luque and Javidan (2004), uncertainty avoidance is "the extent to which ambiguous situations are threatening to individuals, to which rules and order are preferred, and to which uncertainty is tolerated in a society" (p. 602). Societies with high uncertainty avoidance impose more rules and regulations on people, with less tolerance for change and innovation (De Mooij & Hofstede, 2010). In this case, business practices related to the environment will be promoted through the issue of regulations, which will prompt organisations to develop more rigid and standardised actions.
In contrast, societies with low uncertainty avoidance are more receptive to change and have more flexible rules and laws; and thus, stakeholders in these types of societies have higher expectations concerning social and environmental habits and require more information on environmental issues and sustainability in general. This opinion is shared by Adelopo, Cea Moure, and Obalola (2013), who posit that companies from countries with less tolerance to uncertainty disclose more environmental information to reduce uncertainty.
Regarding the positive or negative tendency of the relationships, the various scholars do not agree, reporting mixed results. Scholars such as Vachon (2010) find a negative relationship between uncertainty avoidance, green corporatism, and environmental innovation, while other researchers such as Husted (2005), Orij (2010) and Thanetsunthorn (2015) have not found a positive or negative determinant of the impact of uncertainty avoidance on environmental and social disclosure. Considering the previous arguments, the following hypothesis is proposed: Hypothesis 4: The higher the uncertainty avoidance in a society is, the lower or weaker its corporate environmental disclosure will be.

Long-Term Orientation
Long-term orientation suggests that a society attaches considerable importance to future events and occurrences. Individuals in this kind of society believe that the truth depends heavily on the situation, context and time, and they have a major propensity to save and invest, being known for their astuteness and perseverance . On the other hand, a cultural dimension with a short-term orientation implies that a society gives more importance to the past and the present than to the future.
As cultures with a long-term orientation look to the future, socially responsible investments in companies can lead to sustainable and long-term competitiveness and prosperity. Stakeholders view these cultures as less likely to consider corporate social and environmental investments as an agency cost and a waste of shareholders' resources (Cheng, Ioannou, & Serafeim, 2014). In fact, stakeholders in this type of culture give more importance to reports on social and environmental aspects than to traditional financial reports, as they provide important information about the future. Conversely, in cultures with a short-term orientation, it may be difficult to justify the value of corporate social and environmental investments because their payoffs take time to materialise.
Cultures with a long-term orientation are associated with a social approach, and corporate social and environmental practices are expected to be positively related. Scholars such as Hackert, Krymwiede, Tokle, and Vokurka (2012) have found that investments for preventing pollution and in recycling and waste reduction are carried out primarily by organisations operating in this type of culture. This consideration is consistent with the notion that societies with a long-term cultural dimension are more committed to the preservation of the environment and related sustainability issues. Considering all of the above, the following hypothesis is presented: Hypothesis 5: The greater the long-term orientation in a society is, the higher or stronger its corporate environmental disclosure will be.

Indulgence
This dimension is the latest addition to Hofstede's cultural framework  and is related to gratification versus the control of basic human desires connected with well-being. Indulgent societies are more permissive in relation to the natural human desire for health and well-being, while control societies are more likely to believe that such gratification should be controlled and regulated by strict norms. It is therefore considered unlikely that indulgent societies will sacrifice well-being in favour of the environment. Societies of this kind have some main features, including a perception of personal life control, greater importance of leisure, freedom of speech, and a higher percentage of people declaring themselves to be very happy, while less priority is given to maintaining public order, among other aspects.
As Ismail and Lu (2014) observe, "People in indulgence societies prefer happiness and tend to create a perception of freedom, health, and control over life. Its opposite pole, restraint culture, refers to a society which controls the gratification of the above mentioned desires and feelings" (p. 45).
This cultural dimension therefore concerns the degree to which people seek to regulate their desires and impulses, according to the way they are proposed. Relatively strong control is called restriction, while relatively weak control is called indulgence. Given these definitions, we may posit that organisations within restriction cultures will have more incentives to carry out activities related to the environment, and then make them known to stakeholders. Considering the previous arguments, we propose the following working hypothesis: Hypothesis 6: The more indulgent a society is, the lower or weaker its corporate environmental disclosure will be.

Sample
Our initial sample consisted of 13,178 international firm-year observations from 2004 to 2015. Financial entities were excluded because these companies comply with different rules to those for nonfinancial firms, and therefore, financial statements of these two types of firms are not comparable. From this initial sample, 419 firms were also removed because the data of some of the variables were missing. Thus, the final panel data sample is unbalanced and consists of 12,759 firm-year observations pertaining to 28 countries, which are provided in Table 1. As can be seen, the country with the highest representation is the United States with 28.41%, followed by Japan with 14.02% and the United Kingdom with 9.48%. In contrast to these figures, Portugal represents only 0.23% and Greece is the country with the lowest percentage at 0.08%. All the information was collected from the Thomson Reuters database.
The international companies of the final sample operate within the nine industries shown in Table 2. The sectorial classification used in this research is based on the TRBC economic sector classification by Thomson Reuters. The number of companies from each industry is also provided in Table 2. The sectors with the greatest representation are industrials, consumer cyclical and basic metals with 21.91%, 19.23%, and 13.74%, respectively, and the lowest representation comes from telecommunications services with 3.61%.

Variables
Dependent Variable. The dependent variable, environmental disclosure, is labelled ENVIR_ DISCL. Our environmental disclosure proxy, in line with scholars such as Lee, Kim, Lee, and Li (2012) and Rupp and Mallory (2015), was measured using a multidimensional construct with the purpose of collecting all environmental information reported by the firms of our sample. Thus, our environmental reporting index is measured by the ratio between the unweighted aggregation of 53 items relating to environmental matters presented in Table 3which will take the value 1 if the company reports the item analysed, and 0 otherwise-and the total number of items analysed (53). All the environmental items disclosed by the firms of our sample have been collected from the Thomson Reuters database. These items are included in the Asset ESG Environmental datatypes section. Three areas were explored to construct the environmental reporting index-innovation, resource use and emissions-consistent with Radu and Francoeur (2017) and Wu, Liu, Chin, and Zhu (2018), among others. Thus, our environmental reporting index attempts to respond to questions such as (1) Renewable energy use: Do firms make use of renewable energy? (2) Environmental supply chain management: Do companies use environmental criteria (ISO 14000 or energy consumption, among others) in the selection process of their suppliers or sourcing partners? (3) Emission reduction policy: Do firms have a policy to reduce emissions? (4) Waste reduction total: Do companies report on initiatives to recycle, reduce, reuse, substitute, treat or phase out total waste, hazardous waste or wastewater? (5) Environmental products: Do firms report on at least one product line or service that is designed to have positive effects on the environment or which is environmentally labelled and marketed? (6) Water technologies: Do companies develop products or technologies that are used for water treatment and purification or that improve water use efficiency?
Independent Variables. Cultural issues were measured following the national cultural dimension model created by Hofstede (1980Hofstede ( , 2001, which was enhanced later by . Authors such as Vachon (2010)  The values of the first cultural dimension, power distance (POW_DIST), range from 0 to 100. Scores close to 0 represent a smaller power distance, while scores close to 100 represent a larger power distance. High values (50-100) on the power distance index will account for societies with dictatorships or oligarchies, a smaller middle class, greater income inequality, or political systems changed by revolution, among others, while low values (0-49) are scored by societies with a larger middle class, peaceful conflict resolution, less income equality, or political systems changed by evolution.
The cultural dimension of individualism, INDIV, is expressed with an individualism index score that also ranges from 0 to 100. Scores close to 0 represent societies that tend towards collectivism, while scores close to 100 represent individualist societies. Low values (0-49) account for societies with limited respect for human rights, limited press freedom, a slower pace of life, or older husbands and younger wives, while low values (50-100) account for societies with greater press freedom, smaller age differences between spouses, greater respect for human rights, or a faster pace of life.
Masculinity, MASCUL, is expressed with a masculinity index score that also ranges from 0 to 100. Scores close to 0 represent more feminine societies, while scores close to 100 represent more masculine societies. Low values (0-49) account for societies with fewer people living in poverty, more leisure, longer vacations, or both genders shopping for food, while high values (50-100) account for societies with women who are food shoppers, salary preferred over leisure, more people living in poverty, or more functional illiterates.
Uncertainty avoidance, UNC_AVOID, also ranges from 0 to 100. Scores close to 0 represent societies with weaker uncertainty avoidance, while scores close to 100 represent societies with stronger uncertainty avoidance. Low values (0-49) account for societies with more alcoholism, higher speed limits on motorways, fewer nurses per doctor, or consumers buying more pure and clean products, while high values (50-100) account for societies with less alcoholism, consumers buying more readymade convenience products, or lower speed limits on motorways.
Long-term orientation, LONG_ORIENTATION, also ranges from 0 to 100. Scores close to 0 represent a shorter term orientation, while scores close to 100 represent a longer term orientation. Low values (0-49) account for societies with secondary school students performing poorly at mathematics, small savings, little money for investment, or companies reporting quarterly results, while high values (50-100) account for societies with secondary school students performing well at mathematics, large savings, funds available for investment, or companies seeking market share and long-term profits.
The indulgence versus restraint dimension, INDULG, also ranges from 0 to 100. Scores close to 0 represent a more restrained society, while scores close to 100 represent a more indulgent society. Low values (0-49) account for societies with lower crime rates, a larger police force, lower approval of foreign music and films, or less obesity, while high values (50-100) account for societies with higher crime rates, a smaller police force, freedom of speech being rated as of very high importance, or more obesity.
Summarising, the six cultural dimensions range from 0 to 100, with 50 being the halfway point. Countries with a score under 50 show a low culture score, while 50 or above is considered a high culture score. All the values associated with each cultural dimension are publicly available through the website of Geert Hofstede. Thus, all data for measuring Hofstede's six cultural dimensions for each country have been collected from the Hofstede website. 1 Control Variables. Drawing on past evidence, we take into account several factors that may potentially affect the environmental disclosure index. The legal system is considered as a control variable using two proxies. First, we consider the legal system in which the country operates, labelled as CIVIL_LAW (Kolk & Perego, 2010). Second, we also take into account the efficiency of the judicial system-namely, the degree to which the judicial system of a country can guarantee compliance with laws and recommendations, in line with Vishny (1997, 1998). The efficiency of the judicial system is denoted by EFFICIEN_JUDIC_ SYST and is measured from 0 to 10. The effect of those industries with a high or low impact on stakeholders is also controlled for. In this regard, we categorise industries into critical industries-that is, industries with direct and strong effects on stakeholders-and less critical industries-namely, industries with less impact on stakeholders. This variable is denoted as HIGH_IMPACT_INDUS and is calculated as a dummy variable that will take the value 1 if firms operate in high-impact industries, and 0 otherwise.
Firm size is the fourth control variable considered, defined as SIZE and measured as the logarithm of total assets of companies (Jaggi, Allini, Macchioni, & Zagaria, 2018). Return on assets is also controlled for and was calculated as the operating income before interest and taxes over total assets (e.g., Kim, Park, & Wier, 2012). Leverage, labelled as LEVERAGE, has also been considered as a control variable. It is measured as debt over total assets (Clarkson, Li, & Richardson, 2004;Jaggi et al., 2018). Board size has also been taken into account as a control variable. It is defined as B_SIZE and is measured as the number of board members (Husted & Milton de Sousa-Filho, 2019). Board independence (B_INDEP) is also controlled for, and it is measured as the ratio between the total number of independent directors on the board and the total number of members on the board (e.g., Husted & Milton de Sousa-Filho, 2019). The presence of a CSR committee is also considered as a control variable-labelled as CSR_ COMMITTEE-and is measured as a dummy variable that takes the value 1 if the firm has a CSR committee, and 0 otherwise (Konadu, 2017).
The regional effect has also been controlled for by using five geographic zones: Asia, Europe, Latin America, North America and Oceania. These regions are denoted as ASIA, EUROPE, LATINAMERICA, NORTHAMERICA, and OCEANIA, and they are calculated as dummy variables that will take the value 1 if the country of the sample belongs to the region explored, and 0 otherwise. Another control variable used is the economic growth of a country, which is defined as DEVELOPED and measured as a dummy variable that takes the value 1 if firms operate in a developed country and 0 if firms operate in a developing country (Wei & Wang, 2016).
Finally, year effects (YEAR) are also controlled for, including a set of dummy variables. In Table  4, we offer a summary of all the variables employed in this research.

Methodology
To test our hypotheses, we run the following model: where the subscript "i" represents the firm, "t" refers to the time period, β is the estimated parameter, Ʊ i represents the unobservable time-invariant, firm-specific effects (the unobservable heterogeneity) variable among individuals and constant over time (Greene, 1997), and ϴ it is the disturbance term that varies the cross-time and cross-section joint effect. Firm-specific effects are taken into account to control for firm-particular effects on our dependent variable.

ENVIR_DISCL = + POW_DIST + INDIV + MASCUL
The model is estimated using the dynamic panel data estimator of the generalised method of moments (GMM; Arellano & Bond, 1991;Blundell & Bond, 1998), which lags the dependent variable (introducing the temporal dependency). Contrary to other estimators, the GMM procedure is consistent and efficient because it considers the unobservable heterogeneity (Ʊ i ) by modelling it as an individual effect and by removing it with the first differences of the variables. Additionally, the GMM procedure also takes into account endogeneity and reduces the estimation bias.
The GMM procedure presents the Wald χ 2 test, the Arellano-Bond tests AR(1) and AR (2), and the Hansen test. The Wald χ 2 test shows us the model fitness. Whether a second-order serial correlation in the first difference residuals exists is shown by the Arellano-Bond test AR(2). There will be no second-order serial correlation if the null hypothesis of "no serial correlation" is rejected (p > .1). Additionally, the Hansen test of overidentifying restrictions confirms the Ʊ suitability of the instruments used in the estimation if the null hypothesis of noncorrelation between the instruments and the error term is rejected (p > .1).

Descriptive Statistics
In Table 5, we provide the most important statistics of all the variables used in this research study. Our dependent variable, environmental disclosure index (ENVIR_DISCL), shows, on average, a The ratio between the aggregation of 53 items focused on environmental issues and the total number of items analysed. If the company discloses information concerning each item, it will take the value 1, and 0 otherwise POW_DIST Power distance is one of the six culture dimensions addressed by    Note. Mean, standard deviation and the Shapiro-Francia W test for normality. ENVIR_DISCL is the ratio between the aggregation of 53 items focused on environmental issues and the total number of items analysed. If the company discloses information concerning each item, it will take the value 1, and 0 otherwise; POW_DIST represents the power distance, one of the six culture dimensions addressed by , and ranges from 0 to 100; INDIV represents individualism, one of the six culture dimensions addressed by , and ranges from 0 to 100; MASCUL represents masculinity, one of the six culture dimensions addressed by , and ranges from 0 to 100; UNC_AVOID represents uncertainty avoidance, one of the six culture dimensions addressed by , and ranges from 0 to 100; LONG_ORIENTATION represents long-term orientation, one of the six culture dimensions addressed by , and ranges from 0 to 100; INDULG represents indulgence, one of the six culture dimensions addressed by , and ranges from 0 to 100; CIVIL_LAW is a dummy variable that takes the value 1 if the company operates in a country with civil law, and 0 otherwise; EFFICIEN_ JUDIC_SYST measures the efficiency of the judicial system of a country and ranges from 0 to 10; HIGH_IMPACT_ INDUS is a dummy variable that takes the value 1 if the company operates in an industry with strong and direct environmental impact, and 0 otherwise; SIZE is the log of total assets; ROA is the operating income before interests and taxes over total assets; LEVERAGE is the debt over total assets; B_SIZE is the number of directors on a board; B_INDEP is the proportion of independent directors on boards = Total number of independent directors on boards/Total number of directors on boards; CSR_COMMITTEE is a dummy variable that takes the value 1 if the company has a CSR committee, and 0 otherwise; ASIA is a dummy variable that takes the value 1 if the country is in Asia, and 0 otherwise; EUROPE is a dummy variable that takes the value 1 if the country is in Europe, and 0 otherwise; LATINAMERICA is a dummy variable that takes the value 1 if the country is in Latin America, and 0 otherwise; NORTHAMERICA is a dummy variable that takes the value 1 if the country is in North America, and 0 otherwise; OCEANIA is a dummy variable that takes the value 1 if the country is in Oceania, and 0 otherwise; DEVELOPED is a dummy variable that takes the value 1 if the country is a developed country, and 0 if the country is a developing country.
out of 100, and indulgence (INDULG) is 59.55 out of 100. Furthermore, 42.18% of the firms of our sample operate in a country with civil law (CIVIL_LAW), and the efficiency of the judicial system (EFFICIEN_JUDIC_SYST) is, on average, 9.37 out of 10, which is high. A total of 61.25% of the firms of the sample operate in high-impact industries. Firm size is 9.64 (log of total assets, expressed in euros); return on assets is 6.38%; leverage, on average, is 13.11%; the number of board members is 10.89, with 50.79% of board members being independent; and 59.08% of the firms have a CSR committee. Additionally, 19.80% of the countries in our sample are located in Asia, 32.06% in Europe, 2.88% in Latin America, 38.44% in North America and 6.82% in Oceania, with 91.37% of the firms operating in developed countries. In Table 5, the Shapiro-Francia W test for normality is also provided for each variable. The null hypothesis of this test is that the population is normally distributed (Azat, 2014). Thus, on one hand, if the p value is less than the chosen alpha level, then the null hypothesis is rejected and there is evidence that the tested data are not normally distributed. On the other hand, if the p value is greater than the chosen alpha level, then the null hypothesis that the data came from a normally distributed population cannot be rejected. For all the values of the Shapiro-Francia W test, the p value is >.1, and accordingly, for all the variables, we cannot reject the null hypothesis of normality and we can confirm that our data are normally distributed.
In addition, multicollinearity concerns have been checked by calculating the correlation matrix provided in Table 6. According to the values of Table 6, none of the coefficients is higher than 0.8 (e.g., Pucheta-Martínez, Bel-Oms, & Olcina-Sempere, 2018). Therefore, multicollinearity is not a problem in our analysis.

Multivariate Analysis and Discussion
In Table 7, we present the findings of the nine models built for testing our hypotheses. In Model 1, we explore the association between the cultural dimensions of power distance (POW_DIST) and environmental disclosure. The variable of power distance provides a negative sign, according to our expectations, but it is not statistically significant. Thus, our first hypothesis is not supported, and this finding suggests that power distance does not have an effect on the reporting of environmental information. Our evidence shows that the level of hierarchy in a society is not a determinant factor affecting the disclosure of environmental matters. In other words, a higher or lower power distance in the national culture of each country does not influence the decisionmaking process of firms regarding corporate environmental reporting. This contradicts the work by Waldman et al. (2006) and Peng et al. (2014), who find that managers operating in firms located in countries with a stronger power distance will tend to disclose less environmental information because this cultural dimension induces them to show less commitment to stakeholders' needs. It also contradicts the work by Ho et al. (2012) and Ioannou and Serafeim (2012), who support the thesis that a stronger power distance is positively associated with environmental reporting. Lower environmental disclosure was expected for societies with a large power distance, because in these societies, there will be greater income inequality, more violence in national politics, political systems changing through revolution, a smaller middle class, dictatorships being considered normal, business executives tending to be older, and superiors being considered superior beings and subordinates expecting to be told what to do. In this case, the less powerful members of institutions and organisations will expect and accept that power is distributed unequally, and companies tend to adopt corporate behaviours that are not engaged with stakeholders' needs. A high value for power distance is likely to indicate reduced dialogue between management teams and employees and to mitigate consumer pressure on businesses with regard to environment-related issues such as environmental disclosure. This assumption is supported by Waldman et al. (2006), who suggest that societies with stronger power distance values encourage corporations' managers to show little concern for stakeholders, and firms do  Note. ENVIR_DISCL is the ratio between the aggregation of 53 items focused on environmental issues and the total number of items analysed. If the company discloses information concerning each item, it will take the value 1, and 0 otherwise; POW_DIST represents the power distance, one of the six culture dimensions addressed by , and ranges from 0 to 100; INDIV represents individualism, one of the six culture dimensions addressed by , and ranges from 0 to 100; MASCUL represents masculinity, one of the six culture dimensions addressed by , and ranges from 0 to 100; UNC_AVOID represents uncertainty avoidance, one of the six culture dimensions addressed by , and ranges from 0 to 100; LONG_ORIENTATION represents long-term orientation, one of the six culture dimensions addressed by , and ranges from 0 to 100; INDULG represents indulgence, one of the six culture dimensions addressed by , and ranges from 0 to 100; CIVIL_LAW is a dummy variable that takes the value 1 if the company operates in a country with civil law, and 0 otherwise; EFFICIEN_JUDIC_ SYST measures the efficiency of the judicial system of a country and ranges from 0 to 10; HIGH_IMPACT_INDUS is a dummy variable that takes the value 1 if the company operates in an industry with a strong and direct environmental impact, and 0 otherwise; SIZE is the log of total assets; ROA is the operating income before interests and taxes over total assets; LEVERAGE is the debt over total assets; B_SIZE is the number of directors on a board; B_INDEP is the proportion of independent directors on boards = Total number of independent directors on boards/Total number of directors on boards; CSR_COMMITTEE is a dummy variable that takes the value 1 if the company has a CSR committee, and 0 otherwise; ASIA is a dummy variable that takes the value 1 if the country is in Asia, and 0 otherwise; EUROPE is a dummy variable that takes the value 1 if the country is in Europe, and 0 otherwise; LATINAMERICA is a dummy variable that takes the value 1 if the country is in Latin America, and 0 otherwise; NORTHAMERICA is a dummy variable that takes the value 1 if the country is in North America, and 0 otherwise; OCEANIA is a dummy variable that takes the value 1 if the country is in Oceania, and 0 otherwise; DEVELOPED is a dummy variable that takes the value 1 if the country is a developed country and 0 if the country is a developing country.
*p < .1. **p < .05. ***p < .01.  Note. ENVIR_DISCL is the ratio between the aggregation of 53 items focused on environmental issues and the total number of items analysed. If the company discloses information concerning each item, it will take the value 1, and 0 otherwise; POW_DIST represents the power distance, one of the six culture dimensions addressed by , and ranges from 0 to 100; INDIV represents individualism, one of the six culture dimensions addressed by , and ranges from 0 to 100; MASCUL represents masculinity, one of the six culture dimensions addressed by , and ranges from 0 to 100; UNC_ AVOID represents uncertainty avoidance, one of the six culture dimensions addressed by , and ranges from 0 to 100; LONG_ORIENTATION represents the long-term orientation, one of the six culture dimensions addressed by , and ranges from 0 to 100; INDULG represents indulgence, one of the six culture dimensions addressed by , and ranges from 0 to 100; CIVIL_ LAW is a dummy variable that takes the value 1 if the company operates in a country with civil law country, and 0 otherwise; EFFICIEN_JUDIC_SYST measures the efficiency of the judicial system of a country and ranges from 0 to 10; HIGH_IMPACT_INDUS is a dummy variable that takes the value 1 if the company operates in an industry with a strong and direct environmental impact, and 0 otherwise; SIZE is the log of total assets; ROA is the operating income before interests and taxes over total assets; LEVERAGE is the debt over total assets; B_SIZE is the number of directors on board; B_INDEP is the proportion of independent directors on boards = Total number of independent directors on boards/Total number of directors on boards; CSR_COMMITTEE is a dummy variable that takes the value 1 if the company has a CSR committee, and 0 otherwise; ASIA is a dummy variable that takes the value 1 if the country is in Asia, and 0 otherwise; EUROPE is a dummy variable that takes the value 1 if the country is in Europe, and 0 otherwise; LATINAMERICA is a dummy variable that takes the value 1 if the country is in Latin America, and 0 otherwise; NORTHAMERICA is a dummy variable that takes the value 1 if the country is in North America, and 0 otherwise; OCEANIA is a dummy variable that takes the value 1 if the country is in Oceania, and 0 otherwise; DEVELOPED is a dummy variable that takes the value 1 if the country is a developed country and 0 if the country is a developing country. *p < .1. **p < .05. ***p < .01.
not need to issue more information about environmental matters to achieve the benefits of stakeholders. This leads us to expect a negative relationship between a high power distance and environmental reporting, in contrast to what our evidence has shown. In Models 2, 3 and 6, the effects of individualism (INDIV), masculinity (MASCUL) and indulgence (INDULG), respectively, on environmental disclosure are examined. All variables (INDIV, MASCUL and INDULG) provide a negative sign, as predicted, and are statistically significant. Therefore, Hypotheses 2, 3 and 6 cannot be rejected. These results confirm that the cultural dimensions of individualism, masculinity and indulgence negatively affect the reporting of environmental information. According to Ho et al. (2012), communities in which individualism prevails place importance on independence and freedom, and consequently, this individualism incites people to prioritise individual needs and interests rather than collective demands. This may explain why firms operating in individualist cultures will be less likely to report environmental information, because the sensitivity of the firms' managers towards stakeholders' needs will be lower. According to Akaah (1990), in individualist cultures, workers show behaviours that are less ethical than those shown in collectivist cultures. In this regard, collectivist countries will tend to show greater concern for the effect of business activities on society (Ho et al., 2012) and more sensitivity towards stakeholders' needs (Blodgett et al., 2001). This supports the thesis that collectivist cultures will be more likely than individualist cultures to disclose environmental information, as Garcia-Sánchez et al. (2016) suggest by providing evidence that firms from collectivist contexts will have more incentives to report environmental and social information to their stakeholders because it will improve their decision-making processes. Additionally, it seems that communities where masculinity predominates are less inclined to provide resources, such as environmental information, to their stakeholders. The findings suggest that masculine cultures in our sample show a lower stakeholder orientation, in line with past literature (Peng & Lin, 2009;Roy & Goll, 2014). In masculine communities, where work prevails over family, religion focuses on a powerful God, the strongest are admired, the weakest are disdained, fathers are expected to deal with facts and mothers to deal with feelings and emotional gender roles are distinct (men should be assertive, tough, and focused on material success, while women should focus on the quality of life), companies limit their disclosure of environmental information. In this regard, cooperation, integration, cohesion, agreement, fewer functional illiterates, fewer people living in poverty, both genders shopping for food and more aid given to poorer countries, among others, are not predominant values among companies' managers. It would be expected that these values are more prevalent in feminine contexts and, consequently, would be associated with a higher disclosure of environmental information. Furthermore, indulgent cultures also have a lower tendency to report environmental information, because these communities, in comparison with restraint cultures, support desires such as enjoying life or entertainment (Ismail & Lu, 2014). Members of indulgent communities will also feel healthier and happier, have a perception of personal life control, hold positive and optimistic attitudes, rate freedom of speech for all as very important and display lower moral discipline and ethical standards, inter alia. These feelings are less conducive to the disclosure of environmental information, particularly because indulgent communities display less ethical and moral discipline, and they are less likely than restraint societies to be transparent and require extensive disclosure. In contrast, restraint societies, where moral discipline and ethics are stricter, gratification of needs are supressed and regulated by means of strict social norms, maintaining order of the nation is rated as very important, and personalities are more introverted and pessimistic, ethical and moral discipline are more apparent. Given the characteristics of indulgent cultures, they will have less incentive to satisfy stakeholders' needs and demands through environmental practices.
In Model 4, we analyse the impact of uncertainty avoidance (UNC_AVOID) on environmental disclosure. The coefficient of the variable is positive, in contrast to our predictions, and is significant from a statistical point of view. According to this finding, Hypothesis 4 cannot be supported. Thus, cultures with strong uncertainty avoidance are more likely to report environmental information. Moreover, it also seems that the strict codes of behaviour and beliefs predominant in such communities, as well as the major presence of norms and rules imposed on individuals and the intolerance of unorthodox ideas and behaviour, encourage managers to disclose environmental information. This suggests that more environmentally proactive companies establish norms to ensure certainty and stability. In this regard, Kim and Kim (2009) consider that "CSR-related activities seem to be interpreted by public relations practitioners as one means to guarantee the success of both the organization and society at the same time" (p. 497). Our evidence is not in line with most of the past research on the topic, which shows that strong uncertainty avoidance may result in companies having a weak commitment to sustainability issues such as environmental disclosure (Cordeiro & Sarkis, 1997;Vachon, 2010). Higher disclosure of environmental information is expected in societies with weak uncertainty avoidance, because they maintain a more relaxed attitude in which practice counts more than principles (Nakata & Sivakumar, 1996) and companies' stakeholders often require corporate sustainability practices, such as environmental reporting.
In Model 5, we explore the relationship between the long-term orientation culture dimension (LONG_ORIENTATION) and the reporting of environmental information. The variable exhibits a negative sign, contrary to our predictions, and is statistically significant. Thus, Hypothesis 5 cannot be accepted, and we conclude that communities showing a long-term orientation are less inclined to disclose information related to environmental issues. Although sustainability and environmental issues show their benefits in the long term, our evidence seems to suggest that communities with a long-term orientation are less likely to report environmental information. Environmental reporting is also limited in cultures with a short-term orientation, where virtues related to the past and the present, such as national pride, respect for tradition and fulfilling social obligations, are fostered. Furthermore, firms stress the bottom line and report quarterly results, investors prefer mutual funds and shares, and there is little money for investment and savings, inter alia. Our findings are contrary to the views of Hofstede (2001), who suggests that societies with a long-term orientation will foster pragmatic virtues aimed at future rewards-in particular, perseverance-and will adapt to changing circumstances, and firms will seek long-term profits, there will be funds available for investment and savings, and investors will prefer family businesses and real estate, which would be more consistent with higher environmental disclosure due to its long-term benefits. Our results also contradict the literature on managers' temporal orientation, which would suggest that societies with a long-term orientation emphasise long-term outcomes and priorities, such as benefits or improved performance derived from the disclosure of environmental matters. Thus, our evidence may imply a lower tendency for cultures with a longterm orientation to engage with environmental matters.
Therefore, three out of the six cultural dimensions outlined by Hofstede have a greater contribution to the theory on national culture and environmental disclosure, since their effect on the disclosure of environmental issues contradicts past empirical evidence. In this regard, the findings on power distance report that a higher or lower hierarchical distance or preference for risk and uncertainty do not affect environmental reporting. This leads us to conclude that environmental sustainability is not determined by the law or the power structure that exists to implement it. Concerning uncertainty avoidance, its positive effect shows that firms that are more likely to disclose environmental matters tend to establish rules in order to guarantee certainty and stability. Moreover, practitioners consider environmental activities as a tool for guaranteeing the success of organisations and society. In the case of long-term orientation, and according to our findings, stakeholders give no more importance to reports on social and environmental aspects than to traditional financial reports; and thus, they represent a traditional culture rather than a futureorientated culture. This may be due to the fact that most of the firms in our sample operate in cultures with a short-term orientation, such as Canada, the United States, the United Kingdom and Australia, where traditional cultures prevail.
Our results explain the relationship between national cultures and corporate environmental disclosure and will allow stakeholders to gain an understanding of which cultural dimensions affect environmental disclosure. These results can be useful for managers of multinational corporations, because environmental reporting in different societies requires consideration of national cultures and the social orientation of countries.
Concerning the control variables, the findings show that countries with civil law are less likely to disclose environmental information when individualism and uncertainty avoidance are the cultural dimensions under consideration, but they are more likely to disclose environmental information when long-term orientation is analysed. Furthermore, in all models, except for Model 1, where power distance is explored, a greater efficiency of the judicial system and a larger board size will result in a higher reporting of environmental matters. Firm size is positively associated to environmental information only when uncertainty avoidance is analysed. Additionally, a greater number of independent board members positively affects environmental information disclosure when the masculinity and uncertainty avoidance cultural dimensions are taken into account, and the presence of a CSR committee in a company has a positive effect on the reporting of environmental information in all models, except in Model 1 (power distance) and Model 4 (uncertainty avoidance), where the signs are not statistically significant. The variable of developed countries follows the same pattern as the variable of a CSR committee, but it negatively affects environmental disclosure. Moreover, firms domiciled in Asia show a negative impact on environmental information when the individualism and indulgence cultural dimensions are considered. Firms operating in Latin America show a negative effect on environmental information not only when the individualism and indulgence dimensions are explored but also when uncertainty avoidance is considered. Companies operating in North America only show a negative impact on the disclosure of environmental information when indulgence is analysed. Companies operating in European countries show a positive effect on environmental information disclosure when we take into account the individualism, masculinity and long-term orientation cultural dimensions. The remaining control variables are insignificant. In Table 8, we provide a summary of the expected and obtained signs for each of the hypotheses.

Robustness Analysis
An analysis of robustness was conducted to corroborate our results. In this regard, we used the unweighted aggregation of the 53 items of environmental issues considered for measuring our environmental disclosure index (ENVIR_DISCL) as a dependent variable. This variable ranges from 0 to 53. So as not to extend the article more than necessary, the results of the regressions are not shown here, but they confirm the evidence shown in our baseline models. Therefore, the impact of the cultural dimensions on environmental disclosure is independent of the method used to measure the dependent variable.

Conclusions
Drawing on the diversity of cultures across countries, and specifically using Hofstede's cultural dimensions, the aim of this article was to explore how the cultural context within which firms operate affects their environmental disclosure practices. As proxies for the cultural context, we used Hofstede's six cultural dimensions (normative isomorphism): (1) power distance, (2) individualism, (3) masculinity, (4) uncertainty avoidance, (5) long-term orientation and (6) indulgence.
The findings show that three of Hofstede's six cultural dimensions-individualism, masculinity and indulgence-do indeed have an impact on environmental disclosure, while power distance is not significant, and uncertainty avoidance and long-term orientation are positively and negatively associated, respectively, with the reporting of environmental information, contrary to our predictions. Individualism, masculinity and indulgence have a negative effect on environmental information.
Several implications can be derived from this analysis. First, our evidence confirms that the diversity of cultures across countries affects the reporting of environmental information. Further research could seek to shed some light on the impact of this aspect of the culture approach on other business areas, such as CSR disclosure or firm performance. This evidence may reinforce the theoretical foundations suggesting which factors incentivise firms to disclose environmental issues. Second, our findings provide a solid understanding of which cultural contexts, measured with Hofstede's six cultural dimensions, encourage the reporting of environmental information, and which may be useful for regulatory bodies. Countries with individualist, masculine and indulgent cultures are not the most suitable contexts for disclosing environmental issues. Contrary to our predictions, cultures with a long-term orientation also discourage the reporting of environmental information, while uncertainty avoidance contexts tend to encourage the reporting of environmental matters. Thus, international policy makers might take into account this evidence and recommend or enforce certain aspects over which they may have some influence, such as certain cultural dimensions. At the same time, our evidence can also be very useful for bringing about regulatory homogenisation of environmental disclosure practices with a view to harmonisation throughout the European Union or the United Nations. Third, managers of companies may consider our findings of great interest, and they should pay attention to the disclosure practices on environmental issues in relation to the demands of stakeholders. Stakeholders' expectations should be exceeded by companies, and for this reason, firms should adopt policies and make decisions that are beneficial for society, where culture is a key determinant. In this regard, managers should learn to address cultural differences by implementing proper practices in the cultural context in which they operate; and therefore, they should have a deep knowledge of the country's culture. Our findings provide relevant evidence for managers who are looking to enter new markets with the knowledge needed to learn more about the cultural aspects and who are expecting to be successful in international business. Furthermore, the results of this research may also be relevant for stakeholders, given that they will have more knowledge regarding in which institutional contexts companies are more likely to report environmental information. Finally, our evidence may be useful for other researchers, since this article offers partial empirical evidence of how the diversity of cultures across countries affects environmental disclosure. Our findings disprove some of the main results of past research, which also focused on Hofstede's cultural dimensions. Specifically, contrary to prior empirical evidence, power distance does not affect environmental disclosure, while uncertainty avoidance affects it positively, and long-term orientation affects it negatively. This contradicting evidence signals the need for additional empirical research to better understand the suggested relationship between national cultures and environmental reporting. We hope future researchers can benefit from our findings, and we encourage them to extend our research and to build on these insights. Some future lines of research can be derived from our investigation. We encourage other scholars to extend our research to a sample of companies in both developed and developing countries. It would also be interesting to explore the effects of Hofstede's cultural dimensions on other voluntary disclosures using a sample of financial entities.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors acknowledge financial support from the Spanish Ministry of Economy, Industry and Competitiveness for the research project ECO 2017-82259-R and from the University Jaume I for the research project UJI-B2018-15.