Testing Environmental Kuznets Curve Hypothesis at Firm Level in India

All over the world, deterioration of environmental quality has been considered as a hot issue. The growing number of industries is one of the major reasons for the enhanced level of pollution and fast degradation of environmental resources. This situation urges the corporates to involve themselves in socially responsible activities and to periodically assess the ecofriendly technologies. This paper proposes to investigate the Environmental Kuznets Curve Hypothesis, with reference to sample companies listed in BSE S&P – 500, over the period 1st April 2005 to 31st March 2014. In order to determine the existence of EKC relationship, the sign and magnitude of regression coefficients were analyzed. The results exhibited U shaped relationship for ROE, ROCE, and ROS. At the same time, EKC relationship, by way of inverted U shaped curve, was found for ROA. Based on the empirical findings of this study, some suggestions have been put forward to the corporates and policy makers, to control the level of energy intensity, by implementing eco-friendly technologies.


Introduction
The economic development is an important factor for all countries, in creating and improving the standard of living of people and infrastructure facilities. Many researchers assert that apart from the physical and human capital, the Natural Capital is one of the foremost resources for the development and growth of the economy. Natural Capital is otherwise called as the environmental resources which constitute the main economic assets (Barbier, 2003). In the process of Liberalisation, Privatisation, and Globalisation (LPG), the number of industries has increased all over the world. It is true that industrialization, as a result of LPG, helped to some extend to solve the major economic problems in India by increasing the per capita income, increasing the trade volume, attracting foreign direct investment, generating employment opportunities for the people etc. But any transition in the country's economy also leads to increase in the level of pollution, CO 2 emission, global warming, deforestation etc. (Kander & Lindmark, 2004). These ecological issues were not considered as serious issues until the period of industrial revolution. But after the period of development in industrialization, the level of environment degradation is on the increase. In achieving the growth of industry and to fulfill the demand of the people, the firms have been transformed into mass production process (Kavzoglu, 2008;Wen & Chen, 2008). The production process is mechanized and adoption of technologies depends on the consumption of energy and natural resources like power, oil and gas. The energy and natural resources are as one of the major elements of production. The biological and physical components of industrial production are mixed up with nature and create the environment problems (Dhami et al., 2013). protection activities (Walley & Whitehead, 1994) However, some corporates rightly sensed the fact that enjoying huge level of profit by firms, without the attendant social responsibility, is immoral. Also, if there is no contribution by firms towards the environmental protection, the future profit-earning opportunity of the companies would get affected and they may cease to be sustainable (Trung & Kumar, 2005;Roca, 2003;Porter & Van der Linde, 1995). The different behaviors of the corporates lead to the problem of disequilibrium on the part of companies in respect of their contribution towards the environmental protection. For this reason, it is necessary to maintain the equilibrium among the corporates towards the contribution in ethical activities. The regulatory authorities, corporate decision makers, investors and other stake holders of the company are keenly watching the contributions of companies toward the social responsible activities (Cormier & Magnan, 2003;Stern, 1992). It is very important to understand the nature and relationship between environmental and economic development, before adopting a policy (Coondoo & Dinda, 2002). In the last decades, the core theme of many researches is to examine the relations between the financial performance and environmental performance (Yang et al., 2015;Eroglu & Hofer, 2014;Tiwari et al., 2013). But the study examining the corporate environmental performance is very few.
The present study looks beyond the relationship and examines the existence of Environmental Kuznets curve on the firm level environmental and financial performance in India. The result of this study helps to identify the shape of relationship between environmental and financial performance of the firms. Also, it suggested the place (turning point) where the change occurs in the behavior of corporates towards environmental / social responsibility. Likewise, the findings of this study would help the Socially Responsible Investors (SRI) to understand the social and ethical contribution of the companies. Further, the results of this study would help the policy makers to frame appropriate policy to persuade companies to implement green technology for optimal energy utilization for the production. Hence this research is considered as an important to increase the contribution by companies towards the implementation of ecofriendly technologies for energy savings.

Environmental Kuznets Curve (EKC)
Kuznets Curve is the graphical representation, based on the theory that during the time of economic development, the inequality of the income among people is also increasing. But after attaining a certain level of income, the inequality begins to decrease. This theory was first developed by the great economist and a noble prize winner, Simon Kuznets. Kuznets (1955) found the inverted-U-shaped relationship between per capita income and economic inequality. It explains that increase in per capita income leads to increase in economic inequality. But after the certain stage, increase in per capita income leads to decrease in economic equality. This relationship is represented by a bell shaped curve. This curve is popularly called as the Kuznets Curve.
The Kuznets Curve acquired a new dimension in the later stage of 1990s and onwards. The aspects of environmental degradation and the financial development were examined with the help of Kuznets Hypothesis. The study on the environmental degradation and income could be explained by the inverted-U shaped relationship between the level of environmental degradation and financial development. Grossman and Krueger (1993), Panayotou (1993), Shafik and Bandyopadhyay (1992) discussed the Environmental Kuznets Curve in their research papers. Grossman and Krueger (1993) found the inverted-U shaped relationship between pollution and economic development. Later, Kuznets Curve was extended as the Environmental Kuznets Curve to describe the relationship between environmental quality and income.

Review of Literature
The relevant reviews, relating to the environmental performance and Kuznets Curve, are briefly discussed below to identify the research gaps, estimation techniques and managerial implications. Georgiev and Mihaylov (2015) tested the hypothesis of Environmental Kuznets Curve in countries under the Organization for Economic Cooperation and Development (OECD), with income and air pollutants. The analysis found the existence of environmental Kuznets Curve, with inverted U-shaped relationship between income and pollution. Vasanth et al., (2015) identified the linear and causal relationship between the environmental performance and profitability of the firms in India. Vinayagamoorthi et al., (2015) examined the impact of financial performance on environmental performance of Indian firms and found that three variables, namely, ROA, ROE, and ROS showed significant positive impact but at the same time, one variable, namely, ROCE recorded negative impact on environmental performance of the sample firms in India. Yang et al., (2015) revisited the existence of Environmental Kuznets Curve in China. The study considered the economic performance and emission related indicators such as CO 2 emission, SO 2 emission, industrial waste etc. and could not find evidence of Environmental Kuznets Curve Hypothesis in China. Eroglu and Hofer (2014) investigated the U.S. manufacturing industries about the effect of environmental dynamism on the relationship between inventory leanness and financial performance. The innovative intensity, demand uncertainty and competitive intensity were used as proxy variables to measure the environmental dynamism of the firm. It is found that the innovative intensity helped to increase the effect of inventory leanness on firm performance. At the same time, the competitive intensity and demand uncertainty did not show significant effects. Zhao et al. (2014) examined the energy consumption and efficiency of the manufacturing industries in China and Japan. It is found that in the manufacturing industries of both countries, the level of energy intensity got significantly decreased. However, the structural changes improved the efficiency of the firm. Tiwari et al. (2013) made an attempt to examine the relationship between coal consumption, economic growth, trade openness and CO 2 emissions in India and confirmed the existence of Environmental Kuznets Curve. Duarte et al. (2013) analysed the shape of the relationship between the income and water usage by using regression approach and identified inverted U shape relationship between water usage and income level. Kanjilal and Ghosh (2013) tested the existence of Environmental Kuznets Curve and the relationship between macro level environmental variables and economic performance. This study considered the carbon emission, energy use, economic activity and trade openness in India during the period 1971 -2008. The result of this study identified the existence of Environmental Kuznets Curve in India and found the relationship between carbon emission and per capita income. The study by Saboori and Sulaiman (2013) denied the relevance of Environmental Kuznets Curve in Malaysia and found the bi-directional causal relationship between the economic growth and CO 2 emission. Also, it is suggested that control over the energy consumption may hinder the economic growth. Sagala et al., (2013) tested the Environmental Kuznets Curve and relationship between urbanization and expenditure inequality in Indonesia. The panel data regression was used for the purpose of analysis and found that the result supported the Environmental Kuznets Curve. It is found that the inverted -U shape relationship did exist among the urbanization and expenditure inequality. Mythili and Mukherjee (2011) explored the relationship between river effluents and income of the Indian States. The analysis of this study ignored the existence of Environmental Kuznets Curve and indicated slanting-S shaped relationship. Earnhart and Lizal (2010) investigated the impact of firm level economic performance and environmental performance of Czech firms. The result found that performance of the firm distracted the performance of the firm level environment. Lopez Gamero et al., (2009) examined the relationship between environmental performance, environmental management and performance of the firm by considering the hotels and sample firms in Spain. This study identified that the environmental management system supported the improvement of environmental management and there was positive relationship between the environmental and firms' performance.
It is clear from the review of earlier studies that there was no comprehensive study in examining the existence of Environmental Kuznets Curve in the firm level environmental and financial performance, exclusively in India. The present study takes a step ahead to analyse the existence of Environmental Kuznets Curve in Indian firms.

Conceptual Models
The present study examines the Environmental Kuznets Curve relationship between the environmental performance and financial performance of Indian firms. To test the appearance of curve, many econometric studies were carried out by using quadratic and cubic functions. Normally, these functions reveal the seven different forms of relationship (Kijima et al., 2010;Song et al., 2008;Dinda, 2004). This study also adopts the same quadratic and cubic function and framed four models as follows The above models were administered to identify the different functional forms of relationship between environmental and financial performance of the sample firms and framed the null hypothesis as there is no Environmental Kuznets Curve (EKC) relationship between environmental performance and financial performance of Indian firms. Figure -1 diagrammatically presents the various possible forms of relationship between environmental and financial performance. The line E represents the level of Environmental Pollution while the line Y represents Income / Economic Development (Mythili & Mukherjee, 2011;Kijima et al., 2010;Akbostanci et al. 2009;Song et al., 2008).

Estimation Techniques and Econometric Framework
It is normal to analyse the large companies in the major industries of the economy. In Bombay Stock Exchange of India, many indices are developed based on different criteria. Among the various indices, S&P BSE-500 index reflects the changing pattern of the economy and it considers 500 companies from different major industries. Also this index covers 93% of the total market capitalization of Bombay Stock Exchange. Hence the 500 companies from S&P BSE-500 index were considered for this study but the required data were available only for 191 companies. Hence the final sample size was fixed as 191 companies. The required data relating to Environmental Performance and Financial Performance for a period of 10 years from 1st April 2004 to 31st March 2014, were collected from the CMIE, PROWESS (online corporate database). The other required data were collected from reputed journals and websites.
In order to study the environmental performance of sample firms, parameters such as environmental efficiency score, adoption of ISO 14001, and ratio of toxic wastes were used (Horvathova, 2010). This study adopts the Energy Intensity as a proxy parameter for environmental performance (Vasanth et al., 2015;Vinayagamoorthi et www.ccsenet.org/jsd Journal of Sustainable Development Vol. 8, No. 8;Tung et al. 2014;Sahu & Narayanan, 2011). To measure the financial performance of the sample firms, the present study considered the accounting based measures such as Return on Assets (ROA), Return on Equity (ROE), Return on Capital Employed (ROCE), and Return on Sales (ROS) (Qi, et al., 2014;Iwata & Okada, 2011;Elsayed & Paton, 2005;Hart & Ahuja, 1996).
Further, to find out the existence of EKC relationship between environmental and financial performance, the magnitude of regression coefficient was used (Al-Mulali, et al., 2015;Eroglu & Hofer, 2014;Ogundipe et al., 2014;Sagala et al., 2013;Brajer et al., 2011;Mythili and Mukherjee, 2011). For the parameters (b1, b2, and b3) of these models, homogeneity was assumed (Song et al., 2008). Further, this study applied Descriptive Statistics, and Correlation Coefficient to examine the nature and behavior of the variables and to examine environmental and financial performance of the sample firms in India.

Analysis of the Nature of Environmental and Financial Performance variables
The descriptive statistics describes the nature of environmental and financial performance related variables considered for this study. The descriptive statistics includes mean, standard deviation (SD), minimum, and maximum for the selected variables and the results are summarises in Table -1.
It is to be noted that during the study period, the average value of energy intensity was recorded at 0.0459 (i.e., the value of energy consumption was only 4.59 percent of total sales value). Further, the value of standard deviation for energy intensity was 0.0124. It indicates the fact that the mean value of energy intensity may increase / decrease by 1.24 percent over the value of sale. Minimum and maximum values of energy intensity were 0.0383 and 0.0800 respectively. During the study period, the range of energy intensity moved within the values of 3.83 percent (minimum value) and 8 percent (maximum value) in respect of sample firms.
According to the results of the The results of the Table reveal that the value of energy consumption was only at moderate level (i.e., value of energy consumption was 4.59 percentage over the value of sales). This indicates that the sample firms controlled the level of energy intensity during the study period i.e., the mean value (-0.0459) was only half of the maximum value (-0.0800) of EI. In other words, during the study period, the environmental performance of sample firms increased by way of reducing the level of energy intensity. This clearly shows that there was effective energy consumption by the sample firms during the study period. At the same time, much difference was found in the value of energy intensity (maximum -8 percent and minimum values -3.83 percent). Hence, it is inferred that the sample firms took appropriate steps and control over energy consumption. At the same time,

Source: Compiled from Prowess and computed using SPSS -20
Note: From the results of the Table, it is significant to note that the increase in values for ROA, ROE, and ROCE normally tended to decrease the value of energy intensity (i.e., in nature, the environmental performance variable and financial performance variables recorded inverse relationship and vice versa). At the same time, the financial variable, namely, ROS experienced direct relationship with energy intensity (i.e., increase in the value of ROS (by one unit) leads to increase in the value of energy intensity (by 0.16 unit) and vice versa). This establishes that the sample firms did take adequate measure to improve the environmental performance, along with the financial performance of the firms. The analysis of the Table shows that the increase in the values of ROA, ROE, and ROCE supported the improvement of the environmental performance of the sample firms during the study period. But one variable, namely, ROS leads to environmental degradation by increasing the level of energy intensity. Hence it is suggested t*hat for the environmental performance (to reduce the value of firm level energy intensity) in respect of sample firms, the corporate management should take appropriate steps to increase the level of income corresponding to the value of assets, equity and amount of capital employed by its effective utilization.  Vol. 8, No. 8; In this section, the existence of EKC and the possible shapes of relationship between environmental and financial performance variables was analyzed. For this purpose, the models such as ROA model (equation-1), ROE model (equation-2), ROCE model (equation-3), and ROS model (equation-4)] were tested and the results are summarized as linear function, quadratic function, and cubic function. As mentioned already, these results help to expose the seven different shapes of relationship such as monotonically increasing linear relationship, monotonically decreasing linear relationship, level relationship, U shaped relationship, inverted U shaped relationship (EKC), N shaped relationship and tilted S shaped relationship. According to the Table, linear function of Model -IV shows the value of regression coefficient for ROS was 0.107, along with the constant value at 0.028. Also the magnitude of regression coefficient were -0.965 for ROS (b1), and 2.926 for ROS-squared (b2) of quadratic function. In order to estimate the more possible shapes of relationship, the cubic function was used for Model -IV. It resulted that the coefficient values for ROS, ROS-squared, and ROS-cubed were 0.000, -2.816, and 10.695 respectively. This result indicates that the linear and cubic function does not explicit any shape of relationship between environmental and financial performance variables. But the coefficient values of quadratic function indicates a negative sign (b1 < 0) for ROS and positive sign (b2 > 0) for ROS-squared in respect of sample firms. In short, there was an existence of U shaped relationship between EI and ROS in respect of sample firms.

Analysis of Parameter Estimation for Linear, Quadratic and Cubic Models of Financial Performance Variables with Environmental Performance Variable
The above discussed results indicate that there was an inverted U shaped (i.e., EKC) relationship between ROA and energy intensity. It explains that the consumption of energy was high during the increasing stage of ROA but after the value of turning point, the energy consumption was controlled. At the same time the other financial performance variables such as ROE, ROCE), ROS explicit U shaped relationship in respect of sample firms. The U shaped relationship explains that increase in the value of ROE, ROCE, and ROS supports for energy consumption until the value of turning point is reach. But after the value of turning point the ROE, ROCE, and ROS does not support for environmental performance.
In order to estimate the turning point, coefficient value of quadratic function was used. The turning point values for the above tested models were estimated by -b1/2b2 and constant value (Saboori & Sulaiman, 2013;Song et al., 2008 Table -3 and computed using SPSS -20 The diagrammatical presentation of the shape of relationship between environmental and financial performance variables are presented in Figure 2. In the figure the scale of Y axis stands for the value of energy intensity and X axis represents the value of financial performance variables. Further the figure are classified on the basis of financial performance variables such 2.a, 2.b 2.c, and 2.d for ROA, ROE, ROCE, and ROS respectively. It explains inverted U shaped relationship for that the ROA with energy intensity (2.a), and U shaped relationship for ROE (2.b), ROCE (2.c), and ROS (2.d) with energy. The figure clearly shows that the value of energy intensity increases steadily along with the value of ROA. But after reaching a certain stage, the level of energy intensity was decreased in respect of sample firms. It is important to note that the figure exhibit that the value of energy intensity was in decreasing trend but after touching the value of turning the curve recorded an increasing trend while the value of ROE, ROCE, and ROS also increased correspondingly.

Conclusion and Research Implications
The present study examined the EKC relationship between environmental and financial performance variables such as ROA, ROE, ROCE, and ROS, by using extensive data, covering the period of 1 st April 2005 to 31 st March 2014. First, the results of the summary statistics show that among the selected financial performance variables, ROE performed well during the study period. Pandey (2010), Chandra (1997) mentioned that the value of ROE shows the profitability efficiency of firms over the value of equity. A higher value of ROE implies the effective productivity of the ownership capital. Also it indicates the effectiveness of the sample firm in utilizing its equity funds. Further, the results reveal that the value of energy consumption was at a moderate level. During the study period, the environmental performance increased without increasing the level of energy intensity. This indicates that the sample firms controlled the level of energy intensity during the study period. In other words, there was efficient consumption of energy by the sample firms. Secondly, the correlation matrix summaries the results of linear relationship between environmental and financial performance variables. It is important to note that three variables, namely, ROA, ROE, and ROCE did experience direct relationship with energy intensity while ROS recorded inverse relationship. Third, the estimation of EKC relationship shows that there was existence of EKC relationship between environmental and financial performance of the sample firms. In other words, there was a U shaped relationship for energy intensity with ROE, ROCE, and ROS while there was inverted U shaped relationship with ROA. According to the result of turning points, the sample firms were presently at the developing stage. The result (no EKC relationship exist) of this study is consistent with the result of Mythili and Mukherjee (2011), Akbostanci et al. (2009). At the same time, the findings of this study contradict the results of Ahmed and Long (2012), Stern et al. (1996).
As pointed out by Panayotou (1997), there is a necessity for policy regulations to reduce the environmental degradation at low income level and improve environmental performance at higher income levels. Besides, it is urged that only corporates could initiate appropriate steps to control the level of energy consumption by www.ccsenet.org/jsd Journal of Sustainable Development Vol. 8, No. 8; implementing the eco-friendly technology. Also it is suggested to formulate appropriate policies to control environmental degradation by corporates and to increase the contribution towards the environmental protection activities.
Even though this study has many research implications, the results of this study was interpreted only on the basis of empirical analysis. However, some of limitations are inherent. In this study, the energy intensity was considered as the only variables to measure the environmental performance of the sample firms. This single measure may not produce the same results for the different environmental related variables. Further the results of this study relied only the data used in this study and estimation techniques. The criticisms of the Kuznets curve are also applicable to this study.