TX_1~AT/TX_2~AT International Journal of Energy Economics and Policy | Vol 11 • Issue 6 • 2021170 International Journal of Energy Economics and Policy ISSN: 2146-4553 available at http: www.econjournals.com International Journal of Energy Economics and Policy, 2021, 11(6), 170-179. Is India Financing Its Emissions Through External Debt? Emrah Bese1*, Haven Swint Friday1, Cihan Ozden2 Texas A and M University - RELLIS, Corpus Christi, Texas, USA, 2Near East University, Mersin, Turkey. *Email: phdebese@gmail.com Received: 24 April 2021 Accepted: 05 August 2021 DOI: https://doi.org/10.32479/ijeep.11533 ABSTRACT The main aim of this study is to analyze the effect of external debt on different types of emissions in India as carbon dioxide emissions, methane emissions, emissions from liquid fuel consumption, emissions from solid fuel consumption, and emissions from gaseous fuel consumption. India has a fast growing in external debt especially after 2008 world financial crisis. India has a similar situation to China and Turkey which also started to increase external debt significantly after 2008 world crisis. This study aims to fill the gap in the literature by analyzing the effect of external debt on emissions. This study is the first study in the literature for India. The second aim of the study is to investigate whether inverted U relationship exists between economic development, and carbon oxide emissions, methane emissions, methane emissions, emissions from liquid fuel consumption, emissions from solid fuel consumption, and emissions from gaseous fuel consumption. This study confirmed inverted-U relationship between methane gas emissions and economic development, and emissions from gaseous fuel consumption and economic development. The positive and significant effect of external debt on carbon dioxide emissions, methane emissions, emissions from gaseous fuel consumption and emissions from solid fuel consumption is confirmed by this study. The analysis is important since after 2008 crisis many countries such as China and Turkey besides India started to borrow external debt heavily to create government investments to boost employment market which collapsed due to global economic crisis. This study carries importance since global greenhouse gas emissions may be financed through external debt in India. Since sustainability is the main issue in current world and reduction of emissions is one of the highest priorities of humanity, necessary measures should be taken into account to reduce financing of emissions through external debt in India. This study recommends further analysis to be done with updated intervals. Keywords: External Debt, India, ARDL Model, Emissions, Economic Growth JEL Classifications: Q01, Q56, C01 1. INTRODUCTION This study investigated the effect of external debt (EX) on greenhouse gas emissions which are carbon dioxide emissions (CE), methane emissions (MN), emissions from solid fuel consumption (SFCO), emissions from gaseous fuel consumption (GFCO), and emissions from liquid fuel consumption (LFCO) for India for the period 1971 to 2012. This is the first study in the literature that investigates the effect of external debt on greenhouse gas emissions for India. In this study, CE, MN, EX, GFCO, SFCO and LFCO are chosen to be investigated by Autoregressive Distributed Lag Model by Pesaran et al. (2001). World Bank website is used to gather the data analyzed in this study. Data used in this study are gathered from World Bank database. CE are emissions from manufacturing of cement and burning of fossil fuels. CE are in kilo tons term. MTHN are emissions from industrial methane production and agriculture. MN are in kilo tons of CE equivalent. SFCO are in kilo tons term. SOFC are emissions from mainly of coal use as energy source. LFCO are in kilo tons term. LFCO are emissions from mainly of petroleum derived fuels as energy source. GFCO are in kilo tons term. GFCO are emissions from mainly of natural gas use as energy source. EX is in current US dollars term. EX is the total external debt which includes long-term debt and short-term debt. Economic development (GD) which is gross domestic per capita are in terms of constant 2010 US$. GD2 is the square of GD. Energy consumption (ECM) is in terms of kg of oil equivalent per capita. Time period of this This Journal is licensed under a Creative Commons Attribution 4.0 International License Bese, et al.: Is India Financing Its Emissions Through External Debt? International Journal of Energy Economics and Policy | Vol 11 • Issue 6 • 2021 171 study is between 1971 and 2012 which chosen according to the availability of data in data sources. CE increased by 8.7 times between 1971 and 2012. MN increased by 0.58 times in the same period. SFCO increased by 8.7 times between 1971 and 2012 (Figure 1). GD increased by 2.7 times and ECM increase 1.2 times (Figure 2). EX increased by 41 times and EX increase rapidly after 2008 (Figure 3). GFCO increased 71 times and LFOC increased by 6.84 times between 1971 and 2012 (Figure 1). All the variables used in this study are in increasing trend in India for the period 1971-2012. The main aim of this study is to investigate the effect of EX on emissions so to answer the question whether India is financing its emissions through EX. Although the main of the study is to investigate the effect of EX on emissions, the Kuznets curve relationship is also investigated as second aim of this study in India. There is a gap in the literature for the investigation of the effect of external debt on emissions for India. This study aims to fill this gap by performing this study. H1: EX has a significant impact on CE in India for the period 1971-2012. H2: There is inverted U relationship between CE and GD in India for the period 1971-2012. H3: EX has a significant impact on MN in India for the period 1971-2012. H4: There is inverted U relationship between MN and GD in India for the period 1971-2012. H5: EX has a significant impact on SFCO in India for the period 1971-2012. H6: There is inverted U relationship between SFCO and GD in India for the period 1971-2012. Figure 1: CE, MN, LFCO and SOFC Figure 2: GD and ECM Bese, et al.: Is India Financing Its Emissions Through External Debt? International Journal of Energy Economics and Policy | Vol 11 • Issue 6 • 2021172 H7: EX has a significant impact on LFCO in India for the period 1971-2012. H8: There is inverted U relationship between LFCO and GD in India for the period 1971-2012. H9: EX has a significant impact on GFCO in India for the period 1971-2012. H10: There is inverted U relationship between GFCO and GD in India for the period 1971-2012. Second part discusses literature for the EKC hypothesis and external debt. Third part discusses the methodology used in this study. Fourth part discusses results and discussion of this study. Fifth and final part is conclusion. 2. REVIEW OF LITERATURE In this part, most recent studies in the literature are discussed. Overall summary of this part is given in Table 1. Saxena and Shanker (2017) confirmed negative relationship between external debt and economic development for India for the period 1991 to 2016. Although a negative relationship is found, up to a certain level of external debt, external debt positively affects economic development. Nath (2020) examined the relationship between external debt, export and economic development in India for the period 1970-2018. Nath confirmed that the effect of external debt on economic growth is positive in India. Irfan et al., (2020) analyzed the moderating effect of capital formation for external debt an stock market performance for Pakistan, Sri Lanka, Bangladesh and India for the period 1992-2017. Irfan, Rao, Akbar, and Younis confirmed that capital formation has a positive effect for external debt and stock performance and external debt has negative effect on economic development. Chisti and Shabir (2019) analyzed the effect of external debt on economic development, government spending, revenue, inflation and exports in India for the period 2007-2017 on quarterly data. Chisti and Shabir confirmed that there is no significant relationship Figure 3: EX Table 1: Summary of literature review Study Main findings Saxena and Shanker (2017) Confirmed negative relationship between EX and economic development Nath (2020) Confirmed the positive effect of EX on economic development Irfan et al., (2020) Confirmed the negative effect of EX on economic development Chisti and Shabir (2019) Confirmed no significant relationship between EX and economic development Pahwa (2018) Confirmed the negative effect of EX on economic development Joy and Panda (2019) Confirmed EX negatively affected non-developmental expenditure and positively affected inflation Sinha and Bhatt (2017) Confirmed N-shaped relationship between emissions and economic development Sultan et al., (2021) Confirmed the EKC hypothesis in India Murthy and Gambhir (2018) Confirmed N-shaped relationship between emissions and economic development Khan et al. (2020) Confirmed the EKC hypothesis for panel countries of India, China, and Pakistan Alam and Adil (2019) Did not confirm the EKC hypothesis in India Katircioglu and Celebi (2018) Did not confirm the effect of EX on emissions Beşe et al., (2021) Confirmed the effect of EX on emissions Beşe et al., (2020) Confirmed the coal Kuznets curve Magazzino et al., (2020) Confirmed the coal Kuznets curve Qiao et al., (2019) Confirmed the coal Kuznets curve Shahbaz and Sinha (2019) Recommended new methodologies to be used to investigate the EKC hypothesis Purcel (2020) Recommended new methodologies to be used to investigate the EKC hypothesis Bese, et al.: Is India Financing Its Emissions Through External Debt? International Journal of Energy Economics and Policy | Vol 11 • Issue 6 • 2021 173 between external debt and economic development, external debt and export, external debt and revenue, and external debt and government spending. They found that external debt causes increase in inflation. Pahwa (2018) examined the relationships between external debt, internal debt, population, investment and trade openness for India for the period 1980-2014. Pahwa confirmed that external debt and internal debt affect economic growth significantly and negatively. Joy and Panda (2019) analyzed the relationship between external debt, external debt servicing, gross domestic capital formation, gross domestic savings, developmental expenditure, non- developmental expenditure, export, inflation and foreign direct investment for India. Joy and Panda confirmed the long run relationship between the variables. Joy and Panda confirmed that external debt postively affected inflation but negatively affected non-developmental expenditure. Sinha and Bhatt (2017) analyzed the relationship between nitrogen dioxide emissions (NO2), CE and economic growth for India. Sinha and Bhatt examined CE for 1960-2011 and NO2 for 1970- 2012. Sinha and Bhatt confirmed N-shaped relatinship between emissions and economic growth for NO2 and CE. Sultan et al., (2021) confirmed the EKC hypothesis in India for the period 1978 to 2014. Murthy and Gambhir (2018) confirmed N-shaped relationship between CE and economic development in India for the period 1991-2014. Khan et al. (2020) confirmed the EKC hypothesis for panel countries of China, India and Pakistan for the period 1970-2016. Khan et al. confirmed U-shaped relationship between ecological footprint and economic development for India and China, and the EKC hypothesis for Pakistan for the period 1970-2016. Alam and Adil (2019) did not confirm the EKC hypothesis in India for the period 1971-2016. For the effect of external debt on emissions, the most recent studies are belong to Katircioglu and Celebi (2018) and Beşe et al., (2021). Katircioglu and Celebi (2018) analyzed the case in Turkey and Beşe et al., (2021) analyzed the case in China. While Katircioglu and Celebi (2018) did not find any evidence for the effect of EX on emissions in Turkey, Beşe et al., (2021) confirmed the effect of EX on emissions in China. Since this study analyzed the emissions from coal consumption, the most recent studies for coal consumption and economic development are belong to Beşe et al., (2020), Magazzino et al., (2020) and Qiao et al., (2019). For literature review studies for the EKC hypothesis, the most recent studies are belong to Shahbaz and Sinha (2019) and Purcel (2020). Both literature reviews stated that new methodologies should be included for the investigation of the EKC hypothesis. Literature review shows that the general tendency in the literature is to analyze the relationship between external debt and economic development. There is a gap in the literature for the analysis of the effect of external debt on emissions for India. 3. DATA AND METHODOLOGY 3.1. Data The variables used in this study are as follows. CE are emissions from manufacturing of cement and burning of fossil fuels. CE are in kilo tons term. MTHN are emissions from industrial methane production and agriculture. MN are in kilo tons of CE equivalent. SFCO are in kilo tons term. SOFC are emissions from mainly of coal use as energy source. LFCO are in kilo tons term. LFCO are emissions from mainly of petroleum derived fuels as energy source. GFCO are in kilo tons term. GFCO are emissions from mainly of natural gas use as energy source. EX is in current US dollars term. EX is the total external debt which includes long-term debt and short-term debt. Economic development (GD) which is gross domestic per capita are in terms of constant 2010 US$. GD2 is the square of GD. Energy consumption (ECM) is in terms of kg of oil equivalent per capita. Data used in this study are retrieved from world bank website. Data used in this study range between 1971 and 2012 since the data in world bank website is limited till 2016 for used variables in this study. Range is determined till 2012 for the analyzed variables to provide the stability for established models. 3.2. Methodology ( ) ( ) ( ) ( ) ( ) 2 0 1 2 3t t t t 4 tt ln CE = r +r ln GD + r ln GD + r ln ECM + r ln EX + e (1) ( ) ( ) ( ) ( ) ( ) 2 0 5 6 7t t t t 8 tt ln MN = r +r ln GD + r ln GD + r ln ECM + r ln EX + e (2) ( ) ( ) ( ) ( ) ( ) 2 0 9 10 11t t t t 12 tt ln GFCO = r +r ln GD + r ln GD + r ln ECM + r ln EX + e (3) ( ) ( ) ( ) ( ) ( ) 2 0 13 14 15t t t t 16 tt ln LFCO = r +r ln GD + r ln GD + r ln ECM + r ln EX + e (4) ( ) ( ) ( ) ( ) ( ) 2 0 13 14 15t t t t 16 tt ln SFCO = r +r ln GD + r ln GD + r ln ECM + r ln EX + e (5) Relationship between GD, square of GD, ECM and EX, with CE, MN, GFCO, LFCO and SFCO are modelled above in equation 1-5. For the equation 6 below, v0, v1, v2, v3, v4 are coefficients for the examined variables which are GD, square of GD, EM and EX and bt is for error term. EM is for emissions which are CE, MN, GFCO, LFCO and SFCO. ( ) ( ) ( ) ( ) ( ) 2 0 1 2 3t t t t 4 tt ln EM = v +v ln GD + v ln GD +v ln ECM +v ln EX + b (6) Bese, et al.: Is India Financing Its Emissions Through External Debt? International Journal of Energy Economics and Policy | Vol 11 • Issue 6 • 2021174 The ARDL model which is used as to investigate the relationship between the variables is mentioned as below in equation 7. In the model below, M coefficients are long run coefficients. N coefficients are short run coefficients. bt is for white noise residuals. ( ) ( ) 0 1 1 2 1 2 3 4 1 5 11 2 1 1 2 1 3 1 1 0 0 4 1 5 1 0 0 ln ln b t t t t tt s h c i i i i i i i i i pm i i i i t i i LnEM M M LnEM M LnGD M GD M LnECM M LnEX N LnEM N LnGD N GD N LnECM N LnEX − − − −− − − − = = = − − = = ∆ = + + + + + + + + + + + ∑ ∑ ∑ ∑ ∑ (7) Hypothesis of no cointegration is H0 = M1 = M2 = M3 = M4 = M5 = 0 Hypothesis of cointegration is H1 = M1 ≠ M2 ≠ M3 ≠ M4 ≠ M5 ≠ 0 When cointegration is confirmed, short-run coefficients, long-run coefficients, and error correction model of the ARDL model are as below in equation 8, equation 9 and equation 10. ( ) 0 1 1 2 1 1 0 2 3 4 11 0 0 5 1 0 ln b s h t i t i t i i c m i i tt i i p i t t i LnEM C C LnEM C LnGD C GD C LnECM C LnEX − − = = −− = = − = ++ + + + + = ∑ ∑ ∑ ∑ ∑ (8) ( ) 0 1 2 1 0 2 3 4 0 0 5 1 0 ln b s h t i t i i t i i i c m i i t it i i i p i t i t t i LnEM D D LnEM D LnGD D GD D LnECM D LnEX ECTγ − − = = −− = = − − = ∆ + ∆ + ∆ ∆ ∆ = + + + + + ∑ ∑ ∑ ∑ ∑ (9) ( ) 1 2 1 0 2 3 4 5 0 0 0 ln s h t t i t i i t i i i pc m i i t i i t it i i i i ECT LnEM Y LnEM Y LnGD D GD D LnECM D LnEX − − = = − −− = = = ∆ − ∆ − ∆ − − − ∆ = ∆ ∑ ∑ ∑ ∑ ∑ (10) Equation 8 is to determine the long-run coefficients of ARDL model and equation 9 is to determine the short-run coefficients of ARDL model. Error correction model is specified in equation 10. Zivot and Andrews (1992) unit root test is used to investigate the unit root of the variables in this study. The break dates for the variables are 1997 for CE, 1985 for GD, 1985 for square of GD, 2004 for ECM, 1990 for MTHN, 1982 for GFCO, 2001 for LFCO and 1992 for SOFC. All the variables used in this study are stable at I(1) level. Ramsey Reset Test (RRAT), Heteroskedasticity Test: Breusch- Pagan-Godfrey (BPGODT), Heteroskedasticity Test: ARCH (HTAHT), Heteroskedasticity Test: White (HTWT), Breusch- Godfrey Serial Correlation LM Test (BDGODLMT), Normality Test (NMTT), Cusum Test (CTU) and Cusum square Test (CTUSQ) are used in this study to investigate the stability of each model. 1997 is used as a break for CE-EX relationship. 1990 is used as a break for MN-EX relationship. 1991 is used as a break for GFCO-EX relationship and SFCO-EX relationship. 1991 is used as a break date for LFCO-EX relationship. After carrying out stability tests for each model, bounds test of the ARDL model is applied to decide whether cointegration between the variables exist or not. After cointegration is confirmed, cointegration and long run form is run to investigate the long-run relationships between the variables. Eviews is used for all the calculations in this study. 4. RESULTS AND DISCUSSION Analysis of the relationship between CE-EX, MN-EX, GFCO-EX, LFCO-EX and SFCO-EX in each sub-chapter. ARDL model is used for each relationship and the effect of EX on emissions is analyzed for the period 1971-2012 in India. 4.1. Analysis of CE-EX Relationship Bounds test results show that F-statistics value is 6.69 and it is above I1 value of 1% which is 5.06. Long-run relationship between the analyzed variables is confirmed. For further analysis, ARDL-ECM model is applied to calculate short-run and long-run coefficients for each variable. 1997 is used as structural break in the analysis. Further analysis show that inverted U relationship is not confirmed between CE and GD, and EX has positive and significant effect on CE (Table 3). The long-run relationship between the variables is confirmed with negative coefficient of cointegration equation with significance of 1%. RRAT, BPGODT, HTAHT, HTWT, BDGODLMT and NMTT shows that the model satisfies the criteria for stability (Table 2). Further stability tests are carried out by CTU and CTUSQ tests and the model satisfies the criteria for stability for these tests as well (Figures 4 and 5). This analysis confirms the main aim of the study that EX has positive and significant effect on CE. 4.2. Analysis of MN-EX Relationship Bounds test results show that F-statistics value is 13.17 and it is above I1 value of 1% which is 5.06. Long-run relationship between the analyzed variables is confirmed. For further analysis, ARDL-ECM model is applied to calculate short-run and long-run coefficients for each variable. 1990 is used as structural break in the analysis. Further analysis show that inverted U relationship is confirmed between MN and GD, and EX has positive and significant effect on MN (Table 5). The long-run relationship between the variables is confirmed with negative coefficient of Bese, et al.: Is India Financing Its Emissions Through External Debt? International Journal of Energy Economics and Policy | Vol 11 • Issue 6 • 2021 175 Figure 6: CTU test results for ARDL model of MN-EX relationship Figure 5: CTUSQ test results for ARDL model of CE-EX relationship Table 2: Stability Test Results for ARDL Model of CE-EX Relationship Test F-statistic Jarque-Bera Prob RRAT 0.057861 - 0.8130 BPGODT 0.777274 - 0.7077 HTAHT 0.266431 - 0.8971 HTWT 0.752802 - 0.7305 BDGODLMT 1.469063 - 0.2677 NMTT - 0.897383 0.6384 Table 3: ARDL-ECM test results for CE-EX relationship Variable Coefficient Std. Error t-Statistic Prob. Short-run coefficients D(CE(−1)) −0.433507 0.179387 −2.416608 0.0272 D(CE(−2)) −0.452926 0.142230 −3.184459 0.0054 D(CE(−3)) −0.333486 0.118207 −2.821203 0.0118 D(GD) 8.254545 2.218748 3.720361 0.0017 D(GD(−1)) 7.619568 2.857079 2.666909 0.0163 D(GD2) −0.638575 0.177909 −3.589344 0.0023 D(GD2(−1)) −0.647722 0.230572 −2.809192 0.0121 D(ECM) 1.472394 0.280417 5.250725 0.0001 D(ECM(−1)) 0.953091 0.428490 2.224300 0.0400 D(ECM(−2)) 1.142190 0.332804 3.432015 0.0032 D(EX) 0.063998 0.039061 1.638409 0.1197 D(EX(−1)) −0.071699 0.071397 −1.004225 0.3294 D(EX(−2)) 0.025794 0.080463 0.320565 0.7524 D(EX(−3)) −0.176854 0.061796 −2.861898 0.0108 D(D1997) 0.026637 0.022372 1.190616 0.2502 CointEq(−1) −0.600626 0.150815 −3.982542 0.0010 Long-run coefficients GD 2.468839 2.642847 0.934159 0.3633 GD2 −0.094888 0.220521 −0.430289 0.6724 ECM −1.242166 1.120652 −1.108432 0.2831 EX 0.451679 0.134558 3.356761 0.0037 D1997 0.044348 0.035762 1.240096 0.2318 C −2.349624 10.726289 −0.219053 0.8292 Figure 4: CTU test results for ARDL Model of CE-EX relationship cointegration equation with significance of 1%. RRAT, BPGODT, HTAHT, HTWT, BDGODLMT and NMTT shows that the model satisfies the criteria for stability (Table 4). Further stability tests are carried out by CTU and CTUSQ tests and the model satisfies the criteria for stability for these tests as well (Figures 6 and 7). The analysis shows that there is inverted U relationship between MN and economic growth for India for the period 1971-2012. This analysis confirms the main aim of the study that EX has positive and significant effect on MN. 4.3. Analysis of GFCO-EX Relationship Bounds test results show that F-statistics value is 13.48 and it is above I1 value of 1% which is 5.06. Long-run relationship Table 4: Stability test results for ARDL Model of MN-EX relationship Test F-statistic Jarque-Bera Prob. RRAT 0.346012 - 0.5621 BPGODT 0.956173 - 0.5202 HTAHT 0.595183 - 0.5571 HTWT 0.939207 - 0.5349 BDGODLMT 0.788675 - 0.4669 NMTT - 2.706450 0.2584 between the analyzed variables is confirmed. For further analysis, ARDL-ECM model is applied to calculate short-run and long-run coefficients for each variable. 1991 is used as structural break in the analysis (Table 6). Further analysis show that inverted U relationship is confirmed between GFCO and GP, and EX has Bese, et al.: Is India Financing Its Emissions Through External Debt? International Journal of Energy Economics and Policy | Vol 11 • Issue 6 • 2021176 Figure 8: CTU test results for ARDL Model of GFCO-EX relationship Figure 7: CTUSQ test results for ARDL Model of MN-EX relationship Figure 9: CTUSQ test results for ARDL Model of GFCO-EX relationship Table 5: ARDL-ECM test results for MN-EX relationship Variable Coefficient Std. Error t-Statistic Prob. Short-run coefficients D(MN(−1)) −0.192606 0.094999 −2.027446 0.0539 D(GD) 0.510976 0.717617 0.712045 0.4833 D(GD(−1)) 0.538228 0.773099 0.696195 0.4930 D(GD(−2)) −1.692379 0.693996 −2.438599 0.0225 D(GD2) −0.027690 0.057137 −0.484626 0.6323 D(GD2(−1)) −0.043899 0.062236 −0.705359 0.4874 D(GD2(−2)) 0.129929 0.055772 2.329635 0.0286 D(ECM) 0.221518 0.079634 2.781709 0.0104 D(EX) 0.015420 0.017009 0.906573 0.3736 D(D1990) −0.055863 0.004976 −11.227411 0.0000 CointEq(−1) −0.847836 0.121827 −6.959325 0.0000 Long−run Coefficients GD 3.146752 0.261553 12.031014 0.0000 GD2 −0.235291 0.021665 −10.860388 0.0000 ECM 0.261275 0.094311 2.770358 0.0106 EX 0.063899 0.010036 6.366885 0.0000 D1990 −0.065889 0.009644 −6.831824 0.0000 C −0.408473 1.028746 −0.397059 0.6948 positive and significant effect on GFCO (Table 7). The long-run relationship between the variables is confirmed with negative coefficient of cointegration equation with significance of 1%. RRAT, BPGODT, HTAHT, HTWT, BDGODLMT and NMTT shows that the model satisfies the criteria for stability (Table 5). Further stability tests are carried out by CMM and CMMSQ tests and the model satisfies the criteria for stability for these tests as well (Figures 8 and 9). The analysis shows that there is inverted U relationship between GFCO and economic growth for India for the period 1971-2012. This analysis confirms the main aim of the study that EX has positive and significant effect on GFCO. 4.4. Analysis of LFCO-EX Relationship Bounds test results show that F-statistics value is 10.84 and it is above I1 value of 1% which is 5.06. Long-run relationship between the analyzed variables is confirmed. For further analysis, ARDL-ECM model is applied to calculate short-run and long-run coefficients for each variable. 2001 is used as structural break in the analysis. Further analysis show that inverted U relationship is not confirmed between LFCO and GP, and EX has positive and insignificant effect on LFCO (Table 9). The long-run relationship between the variables is confirmed with negative coefficient of cointegration equation with significance of 5%. RRAT, BPGODT, HTAHT, HTWT, BDGODLMT and NMTT shows that the model satisfies the criteria for stability (Table 8). Further stability tests are carried out by CTU and CTUSQ tests and the model satisfies the criteria for stability for these tests as well (Figures 10 and 11). 4.5. Analysis of SFCO-EX Relationship Bounds test results show that F-statistics value is 4.46 and it is above I1 value of 5% which is 4.01. Long-run relationship Table 6: Stability test results for ARDL Model of GFCO- EX relationship Test F-statistic Jarque-Bera Prob. RRAT 0.273494 - 0.6050 BPGODT 1.017255 - 0.4488 HTAHT 1.003738 - 0.3768 HTWT 0.968455 - 0.4846 BDGODLMT 0.796760 - 0.4607 NMTT - 0.469341 0.7908 Bese, et al.: Is India Financing Its Emissions Through External Debt? International Journal of Energy Economics and Policy | Vol 11 • Issue 6 • 2021 177 Table 7: ARDL-ECM test results for GFCO-EX relationship Variable Coefficient Std. Error t-Statistic Prob. Short-run coefficients D(GFCO(−1)) 0.488247 0.105460 4.629695 0.0001 D(GD) 18.074163 8.526407 2.119786 0.0424 D(GD2) −1.475396 0.680229 −2.168971 0.0381 D(ECM) 5.163414 0.973284 5.305145 0.0000 D(EX) 0.229352 0.096956 2.365528 0.0247 D(D1991) −0.072736 0.060475 −1.202733 0.2385 CointEq(−1) −0.867392 0.095171 −9.114030 0.0000 Long-run Coefficients GD 41.141068 2.550409 16.131166 0.0000 GD2 −3.189567 0.204377 −15.606289 0.0000 ECM 5.952802 1.012707 5.878110 0.0000 EX 0.264416 0.106311 2.487181 0.0187 D1991 −0.083856 0.071675 −1.169948 0.2512 C −164.165369 9.989036 −16.434555 0.0000 Table 8: Stability test results for ARDL model of LFCO- EX relationship Test F-statistic Jarque-Bera Prob. RRAT 2.410850 - 0.1362 BPGODT 1.092169 - 0.4180 HTAHT 1.870781 - 0.1801 HTWT 1.084328 - 0.4240 BDGODLMT 0.671995 - 0.4220 NMTT - 1.199199 0.5490 Table 9: ARDL-ECM test results for LFCO-EX relationship Variable Coefficient Std. Error t-Statistic Prob. Short-run coefficients D(GD) 12.543669 4.404050 2.848212 0.0096 D(GD(−1)) 18.731811 4.818237 3.887690 0.0008 D(GD2) −0.993277 0.353822 −2.807279 0.0106 D(GD2(−1)) −1.560840 0.388453 −4.018089 0.0006 D(ECM) −0.265872 0.828896 −0.320754 0.7516 D(ECM(−1)) 1.236287 0.606389 2.038768 0.0543 D(ECM(−2)) 1.147087 0.592453 1.936165 0.0664 D(ECM(−3)) −1.762657 0.617923 −2.852553 0.0095 D(EX) 0.100969 0.087785 1.150184 0.2630 D(EX(−1)) −0.205482 0.128926 −1.593791 0.1259 D(D2001) −0.199082 0.041550 −4.791439 0.0001 CointEq(−1) −0.300436 0.131110 −2.291477 0.0324 Long-run Coefficients GD −1.685412 11.253471 −0.149768 0.8824 GD2 0.562308 1.083166 0.519134 0.6091 ECM −7.538468 6.546226 −1.151575 0.2624 EX 0.667089 0.505226 1.320377 0.2009 D2001 −0.662643 0.362383 −1.828571 0.0817 C 27.989636 51.982378 0.538445 0.5959 Figure 10: CTU test results for ARDL model of LFCO-EX relationship Figure 11: CTUSQ test results for ARDL model of LFCO-EX relationship between the analyzed variables is confirmed. For further analysis, ARDL-ECM model is applied to calculate short-run and long-run coefficients for each variable. 1991 is used as structural break in the analysis. Further analysis show that inverted U relationship is not confirmed between SFCO and GD, and EX has positive and Table 10: Stability test results for ARDL model of N2- ETLN relationship Test F-statistic Jarque-Bera Prob. RRAT 0.675745 - 0.4207 BPGODT 0.940736 - 0.5426 HTAHT 0.515670 - 0.4775 HTWT 0.957380 - 0.5280 BDGODLMT 3.605245 - 0.0721 NMTT - 0.760169 0.6838 Bese, et al.: Is India Financing Its Emissions Through External Debt? International Journal of Energy Economics and Policy | Vol 11 • Issue 6 • 2021178 Table 11: ARDL-ECM test results for N2-ETLN relationship Variable Coefficient Std. Error t-Statistic Prob. Short-run coefficients D(GD) 3.283821 2.747638 1.195143 0.2454 D(GD(−1)) 2.564799 3.425273 0.748787 0.4623 D(GD(-2)) −7.836568 3.080892 −2.543603 0.0189 D(GD2) −0.242091 0.218190 −1.109543 0.2797 D(GD2(−1)) −0.227386 0.275263 −0.826071 0.4181 D(GD2(−2)) 0.641744 0.244523 2.624474 0.0158 D(ECM) 0.931616 0.333519 2.793292 0.0109 D(ECM(−1)) 0.141965 0.382550 0.371101 0.7143 D(ECM(−2)) 0.065501 0.370397 0.176840 0.8613 D(ECM(−3)) −0.802893 0.396938 −2.022715 0.0560 D(EX) 0.149908 0.032383 4.629278 0.0001 D(D1991) 0.030137 0.020592 1.463546 0.1581 CointEq(−1) −0.618359 0.162180 −3.812789 0.0010 Long-run coefficients GD 2.922026 1.952837 1.496298 0.1495 GD2 −0.210805 0.145462 −1.449213 0.1620 ECM 1.237978 0.456519 2.711775 0.0131 EX 0.242429 0.055891 4.337518 0.0003 D1991 0.048737 0.029651 1.643700 0.1151 C −10.392346 5.953342 −1.745632 0.0955 Figure 12: CTU test results for ARDL model of METH-ETLN relationship Figure 13: CTUSQ test results for ARDL model of METH-ETLN relationship significant effect on SFCO (Table 11). The long-run relationship between the variables is confirmed with negative coefficient of cointegration equation with significance of 1%. RRAT, BPGODT, HTAHT, HTWT, BDGODLMT and NMTT shows that the model satisfies the criteria for stability (Table 10). Further stability tests are carried out by CTU and CTUSQ tests and the model satisfies the criteria for stability for these tests as well (Figures 12 and 13). This analysis confirms the main aim of the study that EX has positive and significant effect on SFCO. 5. CONCLUSION Main findings of this study are as below. 1. No inverted U relationship between CE and GD (Hypothesis 2) 2. EX has significant and positive effect on CE (Hypothesis 1) 3. Inverted U relationship between MN and GD (Hypothesis 4) 4. EX has significant and positive effect on MN (Hypothesis 3) 5. Inverted U relationship between GFCO and GD (Hypothesis 10) 6. EX has significant and positive effect on GFCO (Hypothesis 9) 7. No Inverted U relationship between LFCO and GD (Hypothesis 8) 8. EX has insignificant and positive effect on LFCO (Hypothesis 7) 9. No inverted U relationship between SFCO and GD (Hypothesis 6) 10. EX has significant and positive effect on SFCO (Hypothesis 5). Time period of this study for India is from 1971 to 2012. Period is chosen according to the availability of data on energy consumption side. India’s external debt continued to rise after 2012 till today. The results of this study confirmed hypothesis 1, 3, 4, 5, 9 and 10. The results of this study did not confirm hypothesis 2, 6, 7 and 8. MN Kuznets curve and GFCO Kuznets curve are confirmed by this study. The main aim of this study is to prove the effect of EX on emissions. According to study results, EX has significant and positive effect on SFCO, GFCO, MN and CE. The results of this study confirm that India finances its emissions through external debt. As stated above, although the study period covers from 1971 to 2012, India’s external debt continued to increase after 2012. It is highly likely India continued to finance its emissions through external debt after 2012 till today. Since sustainability is the main issue in current world and reduction of emissions is one of the highest priorities of humanity, necessary measures should be taken into account to reduce financing of emissions through external debt in India. Although this study did not confirm inverted U relationship between SFCO and GD, we recommend further studies to investigate coal Kuznets curve for India which is another gap in the literature and coal is one of the major contributors to greenhouse gas emissions. Beşe et al., (2020), Magazzino et al., (2020) and Qiao et al., (2019) are the most recent studies in the literature that confirmed coal Kuznets curve. The time period analyzed and the country of the study which is India are the limits of this study. This study recommends further policies to be taken to control the use of external debt for creating environmental pollution in India. For future research direction, Bese, et al.: Is India Financing Its Emissions Through External Debt? International Journal of Energy Economics and Policy | Vol 11 • Issue 6 • 2021 179 further studies need to be carried out to analyze the effect of external debt on emissions for developing countries. REFERENCES Alam, R., Adil, M.H. (2019), Validating the environmental Kuznets Curve in India: ARDL bounds testing framework. OPEC Energy Review, 43(3), 277-300. Beşe, E., Friday, H.S., Özden, C. (2020), Coal consumption Environmental Kuznets Curve (EKC) in China and Australia: Evidence from ARDL model. Journal of Applied Business and Economics, 22(11), 25-36. Beşe, E., Friday, H.S., Ozden, C. (2021), The effect of external debt on emissions: Evidence from China. International Journal of Energy Economics and Policy, 11(1), 440-447. Beşe, E., Friday, H.S., Spencer, M. (2021), Analysıs of the relatıonshıp between ıncome growth and, coal consumptıon and emıssıons from nonlınear perspectıve. Journal of Academy of Busıness and Economıcs, 21(1), 52-72. Chisti, K.A., Shabir, T. (2019), Impact of external debt on various macro economic variables: Indian evidence. Journal of Economics, Management and Trade, 25(4), 1-16. Irfan, M., Rao, M.W., Akbar, J., Younis, I. (2020), Impact of external debt on stock market performance and economic growth: Moderating role of capital formation. Journal of Finance and Accounting Research, 2(1), 1-27. Joy, J., Panda, P.K. (2019), Pattern of external debt and its ımpact on macroeconomic variables in India. International Journal of Economic Research, 16(1), 261-272. Katircioglu, S., Celebi, A. (2018), Testing the role of external debt in environmental degradation: Empirical evidence from Turkey. Environmental Science and Pollution Research, 25, 8843-8852. Khan, A., Chenggang, Y., Xue Yi, W., Hussain, J., Sicen, L., Bano, S. (2020), Examining the pollution haven, and environmental Kuznets hypothesis for ecological footprints: An econometric analysis of China, India, and Pakistan. Journal of the Asia Pacific Economy, 26, 462-482. Magazzino, C., Bekun, F.V., Etokakpan, M.U., Uzuner, G. (2020), Modeling the dynamic Nexus among coal consumption, pollutant emissions and real income: Empirical evidence from South Africa. Environmental Science and Pollution Research, 27(8), 8772-8782. Murthy, K.V.B. and Gambhir, S. (2018), Analyzing Environmental Kuznets Curve and pollution haven hypothesis in ındia in the context of domestic and global policy change. Australasian Accounting, Business and Finance Journal, 12(2), 134-156. Nath, S. (2020), An analysis of the relationship among economic growth, external debt and exports in India (1970-2018). Economy, 7(1), 59-68. Pahwa, N. (2018), Impact of debt on short-run and long-run growth: Empirical evidence from İndia. In: Current Issues in Economics and Finance. Germany: Springer. p3-18. Pesaran, M.H., Shin, Y., Smith, R.J. (2001), Bounds testing approaches to the analysis of level relationships. Journal of Applied Econometrics, 16(3), 289-326. Purcel, A.A. (2020), New insights into the environmental Kuznets Curve hypothesis in developing and transition economies: A literature survey. Environmental Economics and Policy Studies, 22(4), 585-631. Qiao, H., Chen, S., Dong, X., Dong, K. (2019), Has China’s coal consumption actually reached its peak? National and regional analysis considering cross-sectional dependence and heterogeneity. Energy Economics, 84, 104509. Saxena, S.P., Shanker, I. (2017), External debt and economic growth in India. Social Science Asia, 4(1), 15-25. Shahbaz, M., Sinha, A. (2019), Environmental Kuznets Curve for CO2 emissions: A literature survey. Journal of Economic Studies, 46(1), 106-168. Sinha, A., Bhatt, M. (2017), Environmental Kuznets Curve for CO2 and nox emissions: A case study of India. European Journal of Sustainable Development, 6(1), 267-276. Sultan, Z.A., Alkhateeb, T.T.Y., Adow, A.H. (2021), Verifying the environmental Kuznets Curve hypothesis in the Case of India. International Journal of Energy Economics and Policy, 11(2), 127-132. Zivot, E., Andrews, D.W.K. (1992), Further evidence on the great crash, the oil-price shock, and the unit-root hypothesis. Journal of Business and Economic Statistics, 10(3), 25-44.