Environmental and socio-economic sustainability in India: evidence from CO₂ emission and economic inequality relationship

ABSTRACT

This study demonstrates the evolution of the relationship between anthropogenic CO₂ emission from fossil fuels use and inequality in consumption expenditure in India based on state-level panel data for the period 1981–2008. Controlling for the scale and composition of economic activities and population, the estimated elasticity of CO₂ emission with respect to economic inequality is found to be equivalent to zero for the overall 28 years period. However, classifying the time frame into pre and post economic liberalization periods (1981–1991 and 1992–2008), reveals that the emission-inequality relationship was insignificant or negative in the pre-liberalization period but turned unambiguously positive and significant in the post-liberalization period. Further classification of the post-liberalization period shows that the positive emission-inequality relationship was statistically insignificant during 1992–1999 and it gained significant strength during 2000–2008. Substantively higher increase in the upper economic strata’s propensity to emit CO₂ made feasible by enhanced access to global markets is discussed as a plausible reason for the positive emission-inequality relationship in the post-liberalization period. The reduced form analysis of the emission-inequality relationship provides a vital policy insight that India has a potential opportunity to harness this synergistic relationship and jointly mitigate the environmental and socio-economic sustainability challenges.

KEYWORDS:

• Scale of carbon-dioxide emission
• economic inequality
• India

JEL CLASSIFICATION CODES:

• Q56
• D63
• O53

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1 According to United Nations 2017 statistics, China ranks first in terms of human population, closely followed by India at the second position; the U.S.A. trails these countries at rank three by an extremely wide margin.

2 In India, as well as globally, the largest source of anthropogenic CO₂ emission is fossil fuel use (Garg and Shukla, 2002; Ghoshal and Bhattacharya, 2012). Therefore, fossil fuel use based CO₂ emission examined here represents the largest subset of the anthropogenic CO₂ emissions. Due to lack of adequate data, other sources of anthropogenic CO₂ emission like deforestation, changes in land use, soil erosion and agriculture are not included in this analysis.

3 Economic inequality is usually measured by income inequality. However, due to lack of disaggregated income data in India, the economic inequality measure used in this study is based on consumption expenditure rather than income. The state level inequality in consumption expenditure was constructed by Das, Sinha & Mitra (2014) based on consumption expenditure survey data from National Sample Survey Organization (NSSO).

4 http://edgar.jrc.ec.europa.eu/overview.php?v=CO2andGHG1970-2016 (accessed on 25th November 2017)

5 https://www.weforum.org/agenda/2017/03/worlds-biggest-economies-in-2017/ (accessed on 25th November 2017)

6 https://en.wikipedia.org/wiki/List_of_countries_by_the_number_of_billionaires (accessed on 25th November 2017)

7 http://www.hindustantimes.com/business/india-s-poverty-rate-lowest-among-countries-with-poor-populations/story-UUrEApaBqRZth6EzRdRK6K.html (accessed on 25th November 2017)

8 According to ‘World Population Prospects: The 2017 Revision’ by United Nations Department of Economic and Social Affairs. https://www.un.org/development/desa/publications/world-population-prospects-the-2017-revision.html

9 A major state Haryana is not included in the analysis due to lack of data on inequality.

10 According to the national estimates of Ghoshal & Bhattacharya (2008), around 68 per cent of CO₂ emission come from solid fuels (mainly coal) and around 24 per cent from petroleum products. They estimate the state level CO₂ emission based on the following fuels: LPG, naphtha, motor gasoline, kerosene, high speed diesel oil, light diesel oil, furnace oil, low sulphur heavy stock, and coal. They acknowledge lack of state level data about the use of the following fuels: natural gas, crude petroleum, lignite, oil shale and peat and also some other secondary fuels like refinery gas, plant condensate, feedstock, jet fuels, briquettes and gas coke. Hence, these are conservative estimates of state level CO₂ emission. Furthermore, this data does not segregate production and consumption based CO₂ emission.

11 Das, Sinha & Mitra (2014) use the state level monthly per capita consumption expenditures, which are weighted averages of rural and urban monthly per capita consumption expenditures. They construct the state level Gini indices by using Atkinson’s (1970) decomposition method. They filled the gaps in the data using the Friedman’s (1962) method of interpolation. They conducted various robustness checks to gauge the performance of the interpolated data.

12 It is important to note that while the data for the state domestic products correspond to fiscal years, the data for CO₂ emission correspond to calendar years. Therefore, the fiscal year data were mapped to the calendar year data with a lag. For example, state domestic product from 1980–1981 fiscal year was used to explain CO₂ emission from 1981 calendar year.

13 Several environmental policies were formulated in India between 1981 and 2008, like the 1986 Environment Protection Act, the 1997 National Environment Appellate Authority Act, the 2006 National Environment Policy. Given the nature of these legislations, one would expect that the effect of these legislations (if any) would be more relevant for anthropogenic CO₂ emissions caused by deforestation, land use change, soil erosion, agriculture, and these legislations are less likely to have direct effect on fossil fuel use. Since the CO₂ emission data used in this study is based on fossil fuel use, instead of the environmental policies, the 1991 economic liberalization policy was chosen as the potential structural break point for the analysis due to its pronounced effect on economic structure of the country which is likely to direct influence the fossil fuel based CO₂ emissions as well as economic inequality. It is also important to note here that the National Renewable Energy Act of 2015 that is specifically targeted towards fossil fuel use is outside the time frame of this analysis due to data limitations.

14 Data source: Databook, Planning Commission, Government of India (2014). http://planningcommission.nic.in/data/datatable/0306/table%203.pdf

http://planningcommission.nic.in/data/datatable/data_2312/DatabookDec2014%20127.pdf

15 The Hausman test assumption is not satisfied by the data, therefore, it could not be used. Instead, the Mundlak test was performed to test the null hypothesis that the time invariant unobservables are not related to the regressors, which implies that the random effects assumption is satisfied.

16 The baseline models were also estimated using Arellano and Bond estimator. Those estimates did not provide any significant evidence of first or second order autocorrelation, and the lagged effects of emission and inequality were not significant either.

17 See F test results for H₀: β_ln(Gini) + β_ln(Gini)*D92t08 = 0.

18 The elasticity of CO₂ emission with respect to Gini index in the post-liberalization period = β_Ln(Gini) + β_Ln(Gini)*D92t08.

19 See F test result for H₀: β_ln(Gini) + β_ln(Gini)*D92t99 = 0.

20 See F test result for H₀: β_ln(Gini) + β_ln(Gini)*D2000t08 = 0.

21 See F test result for H₀: β_ln(Gini)*D92t99 = β_ln(Gini)*D2000t08.

22 The elasticity of CO₂ emission with respect to Gini index in the 2000–2008 period = β_ln(Gini) + β_ln(Gini)*D2000t08.

23 The positive emission-inequality relationship in the post-liberalization period also appears to be hold for different sources of emission data. To check the qualitative robustness of the relationship between emission and inequality, CO₂ emission data from Garg and Shukla (2002) was also tried out. Garg and Shukla (2002) provide state wise CO₂ emission data for India only for two years, 1990 and 1995, and their estimates also include emissions from industries. As a result, their emission magnitudes are larger than those provided by Ghoshal and Bhattacharya (2008, 2012). A panel analysis of the 14 states based on measures of CO₂ emissions provided by Garg and Shukla (2002) has only 28 observations. Therefore, only two of the parsimonious models, (1) and (2), could be estimated. In spite of the limited number of observations, the estimates based on Garg and Shukla (2002) emission data qualitatively align with the key result based on Ghoshal and Bhattacharya (2008, 2012) emission data that the emission-inequality relationship is positive in post-liberalization period. Specifically, in the baseline model (1) the marginal effect of inequality is insignificant; however in model (2), the marginal effect of inequality is positive and statistically significant in the post liberalization period (1995) and the marginal effect of inequality in the pre-liberalization period (1990) is insignificant. Due to the limited sample size, these estimates are not presented in the paper.