Can Increase in the Share of Renewable Energy in Economic Growth Shift Turning Point of EKC?Evidence from Time-series Analysis in India

Purna Chandra Tanti,Illa Sai Srujana,Pradyot Ranjan Jena

National Institute of Technology Karnataka,Surathkal,India

Keywords Economic growth CO2 emission Financial development Manufacturing growth EKC Urban population growth Trade openness

Abstract Global warming has a massive bearing on living beings, the country and environment alike. The primary perpetrator of global warming is carbon dioxide emission. This paper tries to find out the interrelationship between GDP growth and carbon dioxide emissions and the potential of renewable energy use to reduce carbon dioxide emission from industrial activities in the short and long-run. The interrelationship between GDP growth and CO2 emission has traced in this paper.Auto-Regressive Distributive Lag of cointegration model has been used to show the causal relationship among growth variables and CO2 emission. The long-run and short-run relationship of variables have been analyzed by using time series data from 1971 to 2016. This study basically focuses on the existence of EKC in India. One of the objectives of this paper is to find out the existence of renewable energy to reduce CO2 emission in India.The Positive coefficient of renewable energy seen in the findings of the study could be due to the lack of data and also due to the limited renewable resources in India during that period. It has also been noticed that energy consumption and urban population have a negative impact air pollution parameter,thus directing towards higher emission standards and emphasizing the need for transforming fossil fuel-based energy profile into one of renewable energy. The results of the study on trade openness and renewable energy suggest that India should moderate its levels of energy-intensive trade and focus on a trade involving clean energy technologies,which can be sustained only by increasing endogenous renewable energy capacity.

1 Introduction

India’s economy has been growing at an average rate of 6-7%annually since the country embarked on economic liberalization policies in 1991 and in the last Five-year plan(2012 to 2017). This economic growth is associated with energy consumption. When economy grows,the industrial,transport and other allied sectors also grow. The growth of the economy also tends to increase the use of energy. To provide an instance the energy consumption per capita has changed from 268.1 kg to 637.4 kg of oil from 1971 to 2014. Furthermore, CO2emissions per capita have also increased from 0.363 to 1.73 thousand metric tons during this period. The alarming rate at which different energy uses and CO2emission have intensified with economic growth necessitates a look at the interconnection between CO2emission, GDP growth, and the pattern of energy consumption. The main question here whether the developing countries should lay emphasis solely on economic growth, or should go forward with sustainable development policies. The position of the turning point in the EKC hypothesis is determined by energy consumption, economic development and CO2emission. In the Initial stages of a country’s development,it emits more CO2. However,upon reaching a certain stage of development the level of CO2 emissions starts falling. CO2emission can be reduced by increasing the existence of the renewable energy potential and the pollutant reducing technology of the nation. The present paper has examined the presence of EKC within the periphery of India’s renewable energy scenario. An energy-efficient economy requires clean energy to make sustainable development. Appropriate policy decisions in the interest of sustainable development require knowledge of the tradeoffs between the growth of the economy and quality of the environment.

Grossman and Krueger introduced the concept “EKC hypothesis” in 1991. In EKC hypothesis, economic development in the initial stage leads to the detriment of the environment in any country, which is inevitable because, at this stage, the natural resources are exploited and used in different industries. There is a high rate of emission of pollutants during the initial growth phase, and that causes the deterioration of environment.However,after reaching a particular income and growth level,government policies tend to check environmental degradation. The developed economy will use modern and sophisticated technologies to lessen the emission of pollutants.

Although the past literature provides evidence for existence of EKC in India(Mukherjee and Chakraborty,2009;Managi and Jena,2008;Ozturk and Uddin,2012;Kanjilal and Ghosh,2013;Ohlan,2015;Mohapatra and Giri,2015;Sinha and Bhattacharya, 2016;Jena,2018),none of the pieces of this literature have considered the presence of renewable energy as an explanatory variable,which helps to mitigate CO2 emission. (Sugiawan and Managi ,2016) pointed out that in Indonesia, renewable energy has a high potential to mitigate climate change and to bring about reduction in the emission of carbon dioxide. They have estimated the presence of EKC hypothesis in Indonesia from 1971–2010. They have studied the effectiveness of renewable energy in mitigating the emission of CO2. They have used the ARDL approach to co-integration as the estimation method. They have observed that the turning point of EKC was 7729 USD per capita during the period. Electricity generation using renewable sources had a significant impact on the reduction in emission of CO2 in the short and long run.

Against this backdrop, the present paper tries to explore the essence of renewable sources of energy in economic growth and carbon emission in India. Since the last decade, India has been growing as a renewable energy-efficient nation in the world. In March 2019, the cumulative renewable power capacity of India was 78.31 GW.A capacity addition of 42.70 GW grid-connected renewable power was achieved during the last five years from April 2014 to March 2019(MNRE,2019). The country has targeted achieving a cumulative energy capacity of 1,75,000 MW by 2022. India’s position has risen to 4th in wind power installation, 6th in solar power,and 5th in total renewable energy installation to all this development in the renewable energy sector. The Ministry of Renewable Energy has undertaken various projects such as installation of solar street lights, solar home lights, and solar lamps to provide access to renewable energy in rural areas. So,the question that arises is: what is the significant effect of renewable energy in reducing CO2 emission? With this backdrop, the main objectives set for the study are: 1)to study the relationship among economic growth, energy consumption and greenhouse emission in India;2)to examine the manifestation of the role of renewable energy in bringing down CO2 emission of the country.

The paper follows this order: Section 2 focuses on methodology which includes data sources and econometric model specifications. The results and discussion form the econometric analysis explained in Section 3,and the summary and conclusion constitute the last section.

2 Research methodology

2.1 Data source

In the present study, we have used data related to India from 1971 to 2016. This secondary data has been extracted from the World Bank Open Data Source(WDI-World Data Indicator). We have taken CO2 emissions in metric ton per capita,use of energy(EC)in Kg of oil equivalent per capita,GDP growth per capita in current US dollars. Besides, manufacturing growth (annual %), financial development (credit to private sectors as %GDP),trade openness(average of exports and imports as percentage of nominal GDP),urban population growth(annual %) and share of renewable energy (ER) in electricity production in kWh per capita have also been considered as the variables. Renewable energy data has been collected from the statistical review of world energy(http://www.worldbank.org/),and unlike for other variables,data for this variable were available only for the period 1996-2015.

2.2 Model specification

In the model, the variable carbon dioxide emission is specified as CO2, Economic growth as GDP, energy consumption as EC, renewable energy as ER, trade openness as T, manufacturing growth as MFG, financial development as FD and urban population as UP.The effectiveness of the EKC hypothesis can be tested by using Eq. (1),that is,to show the inverted U-shape relationship the CO2emission and GDP growth,β2should be less than zero.

However, there is an alternative model - suggested by Narayan, (2010), and presented in Eq. (2) - for evaluating the EKC hypothesis. In this second model, the long-term and short-term income elasticities are calculated and compared to evaluate the EKC hypothesis. If the long-term income elasticity of CO2 emission is slighter, compared with the short period income elasticity in the long period, it represents the presence of environmental benefits of economic growth so that it can be deemed as confirmation in support of the EKC hypothesis. In the present study,three cases have been estimated,namely,controlled for manufacturing growth and financial development in the first case, controlled for trade openness and urban population in the second case,and controlled for renewable energy,trade openness financial development in the third case.

The ARDL bounds testing method has several advantages over other co-integrating methods since it can be used for any model containing a combination of I (0) and I (1) variables. This ARDL model corrects the presence of correlation between the explanatory variables and the error terms by including appropriate lagged variables.

In this model given below,β,γ,∂, ∋,ρ,ϕandδrepresent coefficients of short run, and represents the coefficients of a long-run multiplier. I (2) variables cannot be used in the ARDL model. So, all the variables have to test for stationarity. Stationarity of the variables has been examined using the Augmented Dickey-Fuller(ADF)unit root test. The optimal lags for individual variables have been selected through the AIC criterion. If the model is co-integrated, it represents the presence of a long-run relationship.

When bounds test results are positive,i.e.,there exists a long-run relationship,two models can be considered:one for the long-run (levels model) and the other for the short-term model (restricted error correction model).The coefficient of the error-correcting term tells about the equilibrating rate at which the short-run coefficients move towards the long-run equilibrium. Post-diagnostic tests have been carried out on the model to check whether it is binding on the assumptions of the ARDL approach. The assumption is that the residuals should be normally distributed, and not serially correlated and homoscedastic; hence, corresponding tests have been conducted. The errors of this model should be serially independent. CUSUM and CUSUM of squares tests examine model stability.

Table 1 Descriptive statistics of variables.

3 Results and discussions

At first,the stationarity of the variables has been tested. Table 1 presents the summary statistics of the variables involved. All variables,except ER,have been considered in their log forms.We have used the log transformation to bring data to normality. The distribution of variables, except for renewable energy (ER), is non-normal, so log has been used to transfer the continuous data into normal. The statistical analysis becomes more valid by transferring into the log. Skewed data has been transferred to log for approximate conformation to normality.The variables have been examined for stationarity using graphical analysis and ADF(Augmented Dickey-Fuller)unit root tests. Graphical analysis results are in Figure1. The graph consists of time plots of all the variables and their first difference. The results of the ADF unit root tests are in Table 2. It is clear from the findings that at their first difference,the variables are stationary,i.e.,all are I(1)variables. As none of the variables is I(2),the assumption of ARDL model stands fulfilled. We can use the ARDL model to test the long-run co-integration between the variables.

In the study,the long-run relationship estimation has been carried out on three models. To allow for a study of the effect of many factors on CO2 emissions, we have used three stable models with combinations of these factors. In the first case,we have controlled financial development and manufacturing growth for the estimation.The second case is controlled for urban population growth and trade openness. The third case is controlled for financial development,trade openness,and renewable energy.

Further, the optimum lag structure has been determined for the three models using the AIC criterion. The top five specifications for each model are in Table 3. Once the lag structures of the models have been finalized,the ARDL bounds test can be undertaken to test the long-run co-integration of variables. There are two critical bound values, I (0) and I (1). When all the variables are stationary at a level, the critical bound value is I (0),and when all are stationary at first difference, the critical bound value is I (1). The variables in the model are integrated of order one. Among the variables, there may be a long-run relationship if only the F-statistic of the bounds test is above the critical bound value at I (1). The bounds test results show that the variables are co-integrated in all the three cases in the long-run,as shown in Table 4.

The coefficients of both long-run and short-run models can be estimated after the bound value. The difference between the observed value and the estimated value, obtained using the long run estimation model, iscalled‘an error correction term’. The coefficient of the lagged term of the error-correction term involved in the short-run model represents the equilibrating speed, i.e. the rate at which the co-integrated variables move to establish long-run equilibrium. Therefore,it makes sense only when the error correction term is between 0 to-1,and is significant. Table 5,signifies the long-run relationship in all the three cases from the bounds test results.

Fig.1 Graphical analysis of variables.

Table 2 Augmented Dickey-Fuller unit root test.

The first and third cases show that Y2 is significant in the long run. The significantly negative coefficient value of GDP square implies the presence of EKC in India (Kankesu Jayantha Kumaran, Sinha and Shahbaz,2018; Verma and Liu,2012). Energy consumption negatively impacts the environment in both long-and shortrun(Farhani and Ozturk,2015;Ozturk and Uddin,2012;Z.Nain,Ahmad and Kamaiah,2017;Kankesu Jayantha Kumaran,Verma and Liu,2012;Sinha and Shahbaz,2018). This finding shows the need for encouraging energyrelated economic activities via clean technology.

Table 3 Model selection summary.

Table 4 Bounds test results.

Table 5 Short-run estimates.

This result can also be inferred from the coefficient of renewable energy(ER)in the third case,significantly positive in the long run, contradictory to our present study. We have taken renewable energy in electricity production in kWh. per capita. It necessarily refers to renewable energy share is very low in the energy profile of India. India’s energy profile is mainly fossil fuel-based. The result indicates that with the current renewable energy technology in place, it is not an effective instrument for reducing environmental pollution in India.Due to the low share of renewables in GDP, it still has not made a large effect on CO2emission. (Pata U.Korkut,2018) has carried out the study and found the same in Turkey. This unexpected positively significant result of renewable energy to the emission CO2is contradictory due to very less share of renewable energy sources till 2015. India has not reached a potential level of renewable energy consumption to mitigate CO2emission. Owing to the causal relationship between consumption of energy and growth of the economy(Ozturk and Uddin,2012; M.Z.Nain,Ahmad and Kamaiah,2017), caution is to be maintained, and renewable energytechnology should introduce potentially. This result shows the unexpected result of the positive coefficient of renewable energy on CO2emissions, so it alarms for a closer examination of the entire lifecycle of the current renewable energy projects in India. Therefore, the government committees on renewable energy production projects should perform a life cycle analysis to provide approvals. The policy suggestion is to increase renewable energy productions with due focus on the establishment of support infrastructure and life cycle of the project.

Table 6 ARDL long-run estimates.

Table 7 Diagnostic tests.

If we see other variables from Cases1 and 3,we realize that financial development doesn’t have any significant impacts on the environment in the long-run. In Case 2,the urban population growth coefficient is positive,implying its adverse effect on the environment. The coefficient of manufacturing sector growth is positive, as expected. This result directs towards a more emission-regulated manufacturing sector. In line with the previous findings of (Tiwari, Shahbaz and Adnan Hye, 2013; Sinha and Shahbaz, 2018; Ahmed and Long, 2012) this result refers to the trade openness: when the model is controlled renewable energy, the long-run coefficient is negative,implying that the trade openness helps in the reduction of CO2 emission through technology transfer.Also,Case 3 shows that the negative impact of energy consumption on the environment is much higher than the positive impact of trade in the long- run. This impact might be due to the kind of energy-intensive trade that India is involved in. Hence, realization of the local renewable energy potential, and involvement clean energy technologies in trade are imperative for the present and the future.

Finally,diagnostic tests have been run on all the three cases. The results of serial correlation,normality,and heteroscedasticity tests are in Table 7. The models have been tested for stability using the CUSUM and CUSUM of squares tests,and the results are presented in Figures 2,3,and 4. These tests indicate that models are stable,and there are no serial correlation,heteroscedasticity, and non-normality in the model.

Fig.2 Stability Tests for Case 1.

Fig.3 Stability Tests for Case 2.

Fig.4 Stability Tests for Case 3.

4 Summary and conclusion

The present study examines the relationship between the growth of the economy in India and the emission of CO2, with the inclusion of the share of renewable energy consumption in industrial activities as an additional covariate in the long-run. The turnaround point has been estimated to be around 2387.09 USD per capita. The estimated turning point is outside the samples; the highest GDP per capita of the sample observed is USD 1605.60 per capita in that point. More research is required to find the existence of a turning point within the sample. In the first stages of economic growth, emission increases due to an increase in fossil fuel energybased production in the economy, which affects the environment negatively. The coefficient of GDP square is negatively significant in the long- run due to high growth in the economy as it tries to change its production technique with new methods and technologies. New methods of production help in lessening emission. So,they will be conducive and beneficial to the environment. This paper has also studied the impact of various factors like financial development,the growth of the manufacturing sector,trade openness,and the urban population on CO2 emission. Evidence has been found for the existence of EKC for CO2 emission in India. Other variables such as energy consumption, manufacturing growth, and urban population growth have a positive coefficient,as they cause more emission of CO2,and so have an adverse impact on the environment. Trade openness can reduce CO2 emission and thereby enhance the quality of the environment.

The model has scale, composition, and technique effects. These effects drive the EKC from the observed impact of the manufacturing sector and energy consumption on the environment. Strong regulations on emission standards have to be implemented to reach the turning point early. Emphasis on India involving itself in clean energy trade is of the utmost importance. India is a vast country with substantial renewable energy potential. The observed positive elasticity of renewable energy on CO2 emission reflects of the ineffectiveness of renewable energy in mitigating the CO2 emission in India. Therefore, projects suggested as alternative cleaner energy sources should undergo critical life cycle analysis

In our study,the coefficient of renewable energy consumption has a positive sign for CO2 emissions. However, there might have been an insufficiency of data and manifestation of renewable energy in India during the period. It also calls for a closer examination of the entire life cycle of renewable energy projects. In this study,we have represented renewable energy (ER) in electricity production in kWh per capita so that it may not be affecting negatively to the CO2 emission. India will be able to mitigate the CO2 emission if we include other renewable energy resources which are abundant across the country. The policy suggestion is to increase renewable energy productions with due focus on the establishment of support infrastructure and life cycle of the projects. In future, the declining cost of renewable energy will make Indian people go for renewable energy sources. The expansion of renewable energy in India will help reduce the use of non-renewable energy sources.Renewable sources of energy have grown stronger, particularly in the power sector. The Government of India is giving importance to solar and hydro- electrical energy for sustainable development. If we reexamine the relation between CO2 emission and renewable energy, we find enough evidence to conclude that there will be chances of negative impact in the days to come.

According to the World Energy Outlook, 2015, there will be a massive difference in the projected energy consumption of 2040 and the present energy consumption because there will be growth in middle-income households and urbanization. So,there should be affordable,reliable,and cleaner energy supply for industrial performance,manufacturing sectors,and domestic use. As the Indian government is campaigning about“Make in India”,the country needs a healthy manufacturing sector with low environmental pollution than the CO2emission can be reduced without interfering with economic growth. This healthy manufacturing needs to be implemented at a suitable pace with caution because energy consumption and economic growth have a causal relationship,given the high cost of transforming India’s energy profile. Thus,higher standards of emission,trade regulations,and improving renewable energy share not only are critical to maintaining sustainable economic growth, but also a sustainable country.