Factors of national environmental performance in sustainability management aspect

The ambitious goals of environmental sustainability stated in international agreements and national programs require developing strategies to achieve them. At the same time, there is a lack of empirical evidence on the environmental performance factors, which can be purposefully changed to achieve an effective result in the short and medium-term. The paper aims to find the institutional factors of national environmental performance, including financial ones, which might be effectively used as environmental sustainability management tools. For this, the relationships between the Environmental Performance Index (EPI), as the dependent variable, and the indicators of control of corruption, the effectiveness of an anti-monopoly policy, financial opportunities, undue influence, corporate culture, innovation output, GDP, and income growth among the poorest population, using a sample of 81 countries, and the technique for constructing nonlinear regression models based on the normalizing transformations for non-Gaussian data were studied.The study findings show that environmental performance can be predicted with sufficient accuracy by a linear model of its dependence on corruption control, minority shareholders protection, judicial independence, favoritism in decisions of government officials, tax incentives, ease of access to loans, and innovation output. Adding GDP per capita to the explanatory variables of the EPI model does not significantly affect the result accuracy but changes the model shape from linear to nonlinear. The paper substantiates ways to apply results for institutional reforms and sustainability management, such as inflation targeting, public credit guarantee schemes, performance-based loans, etc.


INTRODUCTION
A significant part of the problems and global goals of sustainable development are related to the environmental sphere (The United Nations, 2015; European commission, 2019; The White House, 2021). Air pollution, accumulation of plastic waste in the oceans, natural resource depletion, deforestation, desertification, land degradation, loss of biodiversity negatively affect living standards and development prospects. Achieving sustainable development goals in the environmental dimension requires a comprehensive transformation of socio-economic policy and corporate management. The transformation effectiveness depends on the reliability of determining the factors of national environmental performance.
Nowadays, empirical studies based on statistical information and international rankings are currently being used for this. Wendling et al. (2018), in this regard, indicated a modern trend of "data-driven environmental policymaking" growth. However, as follows from the

LITERATURE REVIEW
The literature related to the issue of factors affecting national sustainable development in the environmental dimension has been rapidly expanding lately. The question is to what extent policymakers and managers can use it in practical terms to justify institutional reforms and develop methods for managing environmental sustainability at the macroeconomic and corporate levels. It is worth noting that different environmental indices and indicators became the subject of empirical studies. Some articles focus on individual components of environmental performance associated with climate changes such as carbon dioxide emissions, energy consumption (Povitkina, 2015;Sekrafi & Sghaier, 2018), and urban pollution (Winslow, 2005;Fredriksson & Neumayer, 2013). Another group of empirical studies deals with particular aspects of environmental policy and its consequences. For example, Pellegrini and Gerlagh (2006)  Law & Policy (Yale University) and Center for International Earth Science Information Network (Columbia University). In contrast to absolute indicators, EPI shows how close a country is to establish sustainability targets. It decreases the unwanted effect of national economic structure on environmental performance in empirical studies.
Cross-country comparisons were used on the individual determinants of environmental performance. First, it concerns the impact of corruption (Sekrafi & Sghaier, 2018), development of democracy (Winslow, 2005 It was found that corruption is strongly associated with high emissions of CO2 and environmental degradation. Sekrafi and Sgayer (2018) noted that strengthening control over corruption has a positive effect on the quality of the environment, both directly and indirectly, by reducing the informal sector and, consequently, its energy consumption. Environmental performance is usually also positively associated with democracy, however, as Fredriksson and Neumayer (2013) clarified, if only it comes to the democratic capital stock but not current levels of democracy.
Public governance is usually represented in the empiric studies by the Worldwide Governance Indicators (WGI). These indicators include accountability, political stability, government effectiveness, regulatory quality, and control of corruption. The results demonstrate a positive and significant correlation between WGI as a whole and the national environmental performance. However, conclusions regarding the significance of the effect of each indicator do not coincide (Handoyo & Fitriyah, 2019;Basrija & Handoyo, 2019).
The empirical studies of the relationship between environmental performance and the Gross Domestic Product (GDP) have shown inconsistent results. Іt is often argued that the relationship between the GDP and environmental performance is described by the Environmental Kuznets Curve (Gozgor, 2017;Irdhad & Hussain, 2017;Dkhili, 2019). It has an inverted U-shape, which means environmental degradation with an increase in GDP at an early stage of economic development and a reversal of the trend after reaching a certain threshold.
At the same time, Al-mulali et al. (2015) stated that the environmental Kuznets curve hypothesis does not prove in all countries, but only in upper-middle-and high-income ones. Raymond (2004) underscored the lack of evidence of its validity to developing countries. Dinda (2004) and Fiorino (2011) Lee and Thiel (2017), EPI is irrelevant to GDP growth. Povitkina (2015) substantiated that the environmental Kuznets curve "is not a product of economic forces, but is rather a proxy for other processes occurring in the society", in particular changing the corruption level.
Multivariate models provide an opportunity to explain the ambiguous impact of some other factors on environmental sustainability. For example, Holmberg et al. (2009) and Povitkina (2015) have shown that neither democracy nor high government effectiveness alone is sufficient for sustainable development. Cole et al. (2006) and Sarmidi et al. (2015) proved that the government's degree of corruptibility affects the relationship between foreign direct investment and environmental policy stringency. It was concluded that good governance enhances foreign direct investment inflows and the protection of the environment.
Other independent variables are often used while investigating the determinants of national environmental performance along with the above-mentioned ones. Alhassan et al. (2020) explored the effect of international trade on EPI, combining this variable with indicators of economic growth and government integrity. It was stated that improving the integrity of the government might avert the environmental degradation driven by trade. Dluhopolskyi et al. (2019) used the indicators of population values and beliefs, along with such independent variables as governance effectiveness, democracy index, and the GDP. It was found that the dominance of the self-realization values over the survival ones in the society positively affects EPI.
Another strand of researchers focuses on the impact of human development and human capital on national environmental performance. The human development index (HDI) is a summary measure of life expectancy, education level, and gross national income per capita (UNDP, 2021). Lai and Chen (2020) proved inter-correlation between HDI and EP based on the canonical correlation analysis. Samimi et al. (2011) found that HDI positively and significantly affects environmental performance but not for developing countries. Shahabadi et al. (2017) confirmed this result for selected OPEC countries, supplementing the multivariate model with the other factors (industry value-added, economy openness, carbon dioxide emissions, Internet development, natural resource abundance index, and governance quality). Studying national environmental performance determinants, Kim and Go (2020) included the human capital index (HCI) in the model instead of the human development index. The findings show HCI positively affects EPI primarily through one of its main components -ecosystem vitality.
In addition to avowed environmental sustainability determinants, indicators that characterize national culture, development of infrastructure and tourism, level of urbanization and consumption, population growth, and some others are also widely studied. For example, Almeida and García-Sánchez (2017) indicated a significant effect of transport, infrastructure, consumption of goods, and tourism on environmental performance. Dkhili  The results of empirical studies of factors affecting environmental sustainability lead to the following generalizations. First, it is considered that the indicators dividing national economies according to the degree of economic development (GDP, infrastructure quality, stock market capitalization) are generally recognized determinants of environmental performance. However, links between environmental performance and the factors characterizing the institutional environment in the social dimension remain out of focus.
Secondly, the independent variables in the models, as a rule, represent natural, socio-political, and cultural factors that are relatively stable even from a strategic perspective. In other words, they can hardly be considered environmental management tools, rather the conditions for achieving the development goals. These factors include natural resource abundance, ethnic diversity, democracy, human capital, cultural characteristics, population values and beliefs, etc.
Third, environmental management variables are represented in an overly generalized manner, for example, government effectiveness, reg-ulatory quality, policies, and institutions for environmental sustainability. This level of abstraction complicates the practical use of the empirical studies results by policymakers and managers. Only a few models provide empirical evidence of those environmental performance factors, in particular financial ones, which could be managed for sustainability purposes. Factors like domestic credit to the private sector as a percentage of the GDP (Bernard & Mandal, 2016) and investment in R&D (Dkhili & Dhiab, 2019) are examples. At the same time, the factors, such as the effect of taxation on incentives to invest, ease of access to loans, antimonopoly policy, corporate ethics, protection of minority shareholders, judicial independence, and favoritism in decisions of government officials, can also be considered environmental management tools. These factors are instrumental for managing environmental sustainability because their positive changes are subject to the implementation of short-term and medium-term development programs.

AIMS AND HYPOTHESES
The paper aims to determine the institutional factors of national environmental performance, particularly financial ones, which might be effectively used as environmental sustainability management tools. For this, the following hypotheses were proposed: H1: Factors characterizing the institutional environment in the social dimension significantly affect the Environmental Performance Index.
H2: Environmental performance depends on the effect of taxation on incentives to invest and ease of access to loans.
H3: Environmental performance is positively associated with the innovation output.
H4: National environmental performance can be determined with sufficient accuracy using a linear model of its dependence on the indicators of corruption, innovation output, and instrumental factors for managing environmental sustainability.

METHODOLOGY
Multiple regression analysis provides the opportunity to test the hypotheses of the study. The technique (Prykhodko & Prykhodko, 2021) based on the normalizing transformations and prediction intervals for outlier detection was applied for constructing the multiply nonlinear regression models. This technique consists of four steps. In the first step, a set of multivariate non-Gaussian data is normalized using a normalizing transformation. After that, normalized data are checked for multivariate outliers, and, if ones are detected, they are removed.
In the second step, the nonlinear regression model is constructed based on the normalizing transformation. In the third step, the prediction intervals of nonlinear regression are calculated. Finally, in the fourth step, it is checked whether, among the data for which the nonlinear regression model was constructed, are rows that go beyond the found bounds of the prediction interval of regression. If the outliers are detected, they are removed, and all operations repeat themselves, starting with the first for new data. For detecting outliers in multivariate non-Gaussian data, the study implemented the statistical technique (Prykhodko et al., 2017) based on normalizing transformations and the squared Mahalanobis distance.
The EPI was chosen as the variable under study, taking into account its advantages outlined above. Table 1 shows the independent variables selected to test the formulated hypotheses. X 5 and X 9 are the control variables. If the influence of the GDP is the most significant, then this means that a noticeable increase in EPI cannot be achieved without long-term progress in improving the country's welfare. Another control variable focuses on income growth among the poorest population and thus describes national sustainable development in economic and social dimensions together.

RESULTS
At the first stage of the algorithm application, multivariate data distribution needed checking because well-known statistical methods, for ex-  (Mardia, 1970), which estimates are equal to 21.56 and 95.5 respectively, for the initial dataset indicate the distribution of eight-dimensional data from Table A1 is not Gaussian. The estimate is more than twice the theoretical value of, which is 8.89, and the estimate is almost 20% higher than the theoretical value of, which is 80.
If the criterion based on the squared Mahalanobis distance (Johnson & Wichern, 2007) is still applied to detect outliers in the multivariate non-Gaussian data from   (1) are obtained in the third iteration. For 0.005 significance level, there is only one additional outlier in the data for row 65 (Pakistan). All the removals of the data belong to countries that are the laggards of EPI ranking.
In the fourth iteration for 78 rows of data from Table A1 (except for rows, 42, 65, and 79), the val- To test H4, the six-factor nonlinear regression model based on the Log10 transformation for the seven-dimensional data set of the 81 countries from Table A1 (except for the variable X 5 ) was constructed in the form of model (2). A certain advantage of model (1) with seven factors compared with model (2) with six factors is the smaller widths of the prediction interval. The width of the nonlinear regression prediction interval for model (1) with seven factors is less than for model (2) with six factors for all 79 data rows (with the difference until 26%). The null hypothesis proposes that the observed frequency distribution of the values in model (2) is the same as the Gaussian distribution with zero expectation and 0.041953 standard deviation (there is no difference between the distributions) by the Pearson chi-squared test. Testing sustained the null hypothesis with the 0.05 significance level since the test statistic equals 3.67, less than the critical value from the chi-squared distribution that is 9.49 for 0.05 significance level and 4 degrees of freedom.
The study proved the possibility of constructing the six-factor linear regression model to evaluate the EPI based on the seven-dimensional data set of the 81 countries from Table A1 (except for the variable X 5 ) in the form of model (3).
where ε is a Gaussian random variable defined residuals,  Table AI (except for row 42). That is, for 0.005 significance level, there is one outlier in the data for South Africa. It can be explained by a complex of reasons. First, there is a significant difference between issue categories ranks due to the energy consumption structure. South Africa's main environmental problem is a high level of emission of greenhouse gases and pollutants. Around 80% of its emissions come from the energy supply, mostly provided by coal-fired power plants (The Carbon Brief, 2021). As a result, South Africa is one of the laggards in Air Quality, ranking 166th out of 180 countries. At the same time, the country ranks first in terms of Marine Protected Areas. Besides, some predictors from initial data were in the range of values typical of the countries with the better EPI ranks. Given that the effect of institutional reforms manifests itself with a certain lag in time, it is quite indicative that the country climbed up 47 notches in the EPI 2020 ranking.
The null hypothesis proposes that the observed frequency distribution of the ε values in model (3) is the same as the Gaussian distribution with zero expectation and 5.39362 standard deviation by the Pearson chi-squared test. Testing sustained the null hypothesis with the 0.05 significance level since the test statistic equals 5.82, less than the critical value from the chi-squared distribution that is 9.49 for 0.05 significance level and 4 degrees of freedom.
The well-known prediction accuracy metrics are used to judge the prediction accuracy of regression models. These metrics include a multiple coefficient of determination R 2 , a mean magnitude of relative error MMRE, and prediction percentage at the level of magnitude of relative error (MRE) of 0.25, PRED (0.25). The R 2 , MMRE, and PRED (0.25) values equal respectively 0.8183, 0.0576, and 1 for the seven-factor nonlinear regression model (1) with the estimators of parameters, which are calculated for the 78 data rows. The R 2 , MMRE, and PRED(0.25) values equal respectively 0.7529, 0.0796, and 0.9750 for the six-factor linear regression model (3) for the 80 data rows, and equal respectively 0.7317, 0.0785, and 0.9747 for the six-factor nonlinear regression model (2) for the 79 data rows. The R 2 and MMRE values are somewhat better for the seven-factor nonlinear regression model (1) in comparison with the above six-factor regression models, both nonlinear (2) and linear one (3). However, a linear regression model makes it possible to interpret the regression coefficients as the expected change in EPI for a one-unit change in an independent variable when all the others are held fixed.
Model (3) demonstrates that independent variables X 4 , X 6 , and X 7 , which describe judicial independence, favoritism in decisions of government officials, protection of minority shareholders, and control of corruption, essentially affect EPI. In other words, the study results fully support H1 about the significant relationship between factors characterizing the institutional environment in the social dimension and environmental performance. Model (3) proves that EPI is significantly dependent on financial factors (the effect of taxation on incentives to invest and the ease of access to loans). This fact confirms the validity of H2.
The presence of the innovation output factor in model (3) correlates to a certain extent with the study results of Dkhili and Dhiab (2019) regarding the effect of investment in R&D on environmen-tal performance. A change in the innovation output sub-index affects EPI less than a change in any other independent variable in the model. This suggests that national economy innovativeness does not guarantee green technologies' implementation if there are no effective sustainable development incentives. Thus, the study only partially supports H3.
The linear six-factor model (3), built without the factor of the GDP, turned out to be comparable in prediction accuracy with the seven-factor one, which includes this factor. In addition, unlike models (1) and (2), it applies to all countries in the EPI rating except for South Africa, which, as shown above, has a logical explanation. The model fully supports H4 about the possibility of improving EPI in the medium or even short term using environmental management tools and controlling corruption.

DISCUSSION
The positive impact of control of corruption on environmental sustainability is consistent with empirical evidence of Pellegrini and Gerlagh (2006), Sekrafi and Sghaier (2018), Handoyo and Fitriyah (2019), etc. The relationship between EPI and such a variable as Undue influence (the proxy for indicators of judicial independence and favoritism in decisions of government officials) is quite understandable. Preferential treatment of certain companies or oligarchic groups by officials in the process of monitoring air pollution emissions and water contamination, environmental permitting, and decision making on state support for corporate sustainability programs harms incentives for eco-friendly activities. Lack of judiciary independence has a similar negative effect when considering lawsuits regarding changes in environmental legislation and when challenging penalties against companies for violating environmental standards.
Paradoxical at first glance is the high value of the regression coefficient before the variable X 6 (minority shareholders protection). However, this can be explained by the fact that the indicator of minority shareholders' protection is the only variable in the model that characterizes a company's relationship with vulnerable stakeholder groups, its compliance with the principles of corporate social responsibility. In the countries where companies act in socially responsible ways towards minority shareholders, especially if the CSR rules are enshrined in government and industrial regulation, they would likely act in the same manner towards employer associations, local communities, and other stakeholders, based on the balance of economic, social and environmental dimensions of sustainable development. For managing this factor of environmental performance, its significant dependence on the dividend policy and regulation of the squeeze-out procedure is important (Rohov et al., 2020).
The combination of such factors as taxation on incentives to invest and the ease of access to loans in model (3) characterizes the opportunities to attract financial resources and the conditions for their investment, including in green technologies. It is worth noting that as components of the 8 th pillar of the Global Competitiveness Index (financial market development), characterizing access to financial resources, ease of access to loans, and financing through the local equity market are used. The latter component shows to what extent the firms can raise money by issuing shares or bonds on the national capital market. This indicator correlates with the variables X 2 and X 3 used in the model, since the availability of borrowed funds in the form of bank loans and bond issues is usually determined by similar factors, and the opportunities for public equity offerings and equity private placements depends on corporate tax incentives. However, the effect of taxation on incentives to invest has an obvious advantage in the context of the subject of the study, since it, in contrast to the indicator of financing through the local equity market, also characterizes incentives for real investment, including foreign investment.
The significant impact of the financial factors on EPI leaves open questions on how to apply them as instruments in environmental management. The idea that low tax rates always have a positive effect on investment, while high tax rates, on the contrary, is rather simplified. In the medium-term, low tax rates can increase the budget deficit to a critical level. Under these circumstances, government borrowings divert companies' financial resources from investing in new technologies and equipment.
In practice, various forms of targeted tax incentives are widely used to enhance the effect of taxation on investments, for example, reduced tax rates, investment allowances, tax credits, accelerated depreciation, and tax holidays. The targeted tax incentives can cause unjustified disparities in investment directions. However, their use is advisable to stimulate sustainable development since the social and ecological externalities of production activity do not receive an adequate economic assessment in the conditions of the existing free market. Some countries at the top of the EPI rankings applying investment allowances and accelerated depreciation take into account the environmental component of the investment effect. In particular, companies in Australia can claim an immediate deduction for expenditure on activities undertaken to prevent environmental pollution and to treat, clean up, and store waste (Australian Taxation Office, 2021). In the Netherlands, accelerated depreciation applies only for several specific assets, first of all, which are in the interest of environmental protection (PwC, 2018).
There is practical evidence that tax incentives can be quite effective. For instance, Jun (2010), studying Korea's tax system development, emphasized that they play a role in preventing companies from moving into the informal sector. This is important in the context of the EPI factors research since the shadow economy is not associated with environmental sustainability. Bermperoglou et al. (2019) proved the positive effect of investment allowances and tax credits on output, which materializes in two or three years. At the same time, the decision to use tax incentives, like any financial instrument, requires a comprehensive analysis, taking into account the specifics of the national economy and investment climate. It would be appropriate when drafting a tax incentive scheme in a developing country to consider a list of items developed by the United Nations Department of Economic and Social Affairs (UN-DESA) and the Inter-American Center of Tax Administrations (UN-DESA, 2018). In environmental management, finding the optimal solution that balances environmental tax increase, required to reduce the harmful effects of economic activities on the one hand and reduce a tax burden on enterprises through investment incentives on the other one, is of great importance.
Waiver of declared incentives and the deterioration of taxation conditions have an extremely negative impact on the investment climate and prospects for sustainable development. Indicative, in this regard, is the tough reaction of the investment market to the cutting the green tariffs by the Ukrainian government and to the bill on introducing an excise tax on renewable energy. Nordic Environment Finance Corporation (NEFCO), in particular, officially announced its decision "to stop investing in new projects in the renewable energy sector in Ukraine until a predictable and sustainable framework is in place and agreements are being honored by the Ukrainian government" (NEFCO, 2021). The effect of taxation on incentives to invest is leveled out in conditions favorable for capital transfer to low or no-tax jurisdictions. National implementation of Action Plan on Base Erosion and Profit Shifting (OECD, 2013) contributes to the solution of this problem.
Another financial factor that affects EPI, as follows from model (3), shows how easy businesses can obtain a bank loan. First, it depends on the overall economic situation in the world, country and industry, expected trends, and enterprise creditworthiness. The influence of corruption is also emphasized (Qi & Ongena, 2019;Liu et al., 2020). It is important to stress that managing these factors might lead to a significant increase in funding for corporate environmental programs in the medium and even short term. This challenge is especially relevant for developing countries where firms' access to bank loans is difficult. However, during economic crises and nowadays, owing to the spread of the coronavirus pandemic, the problem of obtaining loans exists in highly developed countries too. For instance, the survey on the access to finance of enterprises in the euro area in the period from October 2020 to March 2021, conducted by the European Central Bank (2021), indicates no significant improvement in the availability of external financing and demonstrates faster growth rates of demand for external funds than availability.
In emerging and developing economies, where inflation targeting is not practiced, national currency depreciation, as pointed by Ha et al. (2019), has large and persistent effects on inflation. If under these circumstances, the government does not im-plement a policy for tying up money, particularly in infrastructure projects, the central bank usually raises the discount rate, which increases the cost of borrowed capital. Financing green technologies requires the ease of access to long-term loans. In weak economies, banks associate long-term lending with high risks, which significantly complicates fixed investment. To a certain extent, these risks are explained by the poor financial management of the borrowing enterprises. In favorable market conditions, such enterprises unjustifiably increase the debt-to-equity ratio, which leads to problems with the payment of interest and loan repayment in the event of declining sales.
To support developing country policymakers in solving the problem of improving access to finance, as a key component of entrepreneurship policy framework, UNCTAD has developed a complex of recommended actions. It includes the development of public credit guarantee schemes, incentives for creating private mutual guarantees, promotion of foreign direct investment and collateral-free loan screening mechanisms, providing performance-based loans and stimuli for innovation and green growth, facilitating the use of the intellectual property as collateral, etc. (UNCTAD, 2021). The topicality of these measures is growing against the backdrop of fears of a sudden capital outflow from emerging and developing economies in the event of Taper Tantrum repeat.
Confirmation of H4 indicating that a high GDP level does not belong to the fundamental premise of improving environmental standards is consistent with the results of some other empirical studies. For instance, Pellegrini and Gerlagh (2006) argued that environmental standards are more likely to reflect institutional quality than income levels. Dluhopolskyi et al. (2019) indicated the absence of a fatal character in the relationship between economic growth and national environmental performance. It was underscored that investments in the quality of management and institutions more significantly affect EPI.

CONCLUSION
This paper used the international rankings datasets, the World Bank statistics, the sample of 81 countries, and the technique for constructing nonlinear regression models based on the normalizing transformations for non-Gaussian data. It was revealed that environmental performance is predicted with sufficient accuracy by a model of its dependence on the indicators of corruption, innovation output, and the factors, which can be considered like environmental management tools. The latter is instrumental for managing environmental sustainability because their changes are subject to short-and medium-term development programs. These include the effect of taxation on incentives to invest, ease of access to loans, minority shareholders protection, judicial independence, and favoritism in decisions of government officials.
GDP per capita cannot be regarded as the main determinant of environmental performance. Including this indicator in the EPI model, together with the above factors, does not significantly affect the result accuracy but changes the model shape from linear to nonlinear. At the same time, the study makes it possible to assert that EPI depends on the institutional environment in the social dimension, access to financial resources, and economic innovativeness.
The findings may be of interest to policymakers and practitioners. Reforms to strengthen judicial independence and prevent favoritism in decisions of government officials are of great importance. This is a prerequisite for the adoption of effective environmental laws and their proper implementation. A positive result can be expected from the inclusion of the principles of corporate social responsibility in government and industries regulation.
One of the financial policy's key objectives is finding a balance between an environmental tax increase and a reduction of the general tax burden on enterprises through investment incentives, such as allow-ances, accelerated depreciation, tax credits, and holidays. When solving this problem, the national economy specifics, investment climate, and risks to deteriorate the taxation conditions should be considered. Since the effect of tax incentives diminishes in conditions favorable for capital transfer to low or no-tax jurisdictions, the Action Plan on Base Erosion and Profit Shifting implementation on a national level is advisable.
For sufficient funding state, regional, and corporate environmental programs, the financial policy has to include inflation targeting and tying up money in infrastructure projects. The multi-action approach comprising public credit guarantee schemes, incentives for creating private mutual guarantees, collateral-free loan screening mechanisms, performance-based loans, stimuli for innovation and green growth, and intellectual property used as collateral could be effective for developing countries. As the access to loans depends on firm creditworthiness, environmental sustainability strategies also require financial management improvement.
The prospect for further research is to determine the effect of specific reforms in the financial, tax, investment policies, the judicial system, and corruption control on the EPI score. The studies have to compare environmental performance in both national economies and groups of countries with a common socio-economic policy, such as the EU.