Energy Efficiency Versus Consumption: An Empirical Investigation of Jevons’ Paradox Across Sectors in India

Abstract

Modernisation theorists reiterate endogenous growth models by acknowledging technological production for a rise in the efficiency of energy-use. Yet, world system analysts blame economic production for causing environmental degradation. As economic and demographic modernisation continue to raise the magnitude of extraction despite their declining economic intensity, the dichotomy draws attention to Jevons’ paradox. Debates ingeminating ‘The End of Cheap Oil’ raise arguments about India’s energy deficit despite measures to raise its energy self-sufficiency. Do policies aiming at reducing energy consumption through efficiency enhancement fulfil their targets or does the reality reverse the objectives of the policy? Focusing on the two pillars of energy policy (efficiency-demand and supply), Ordinary Least Square (OLS) technique and Granger causality have been employed to model the energy efficiency–consumption nexus and other determinants of energy consumption, across sectors. The magnitude of energy intensity effect is stronger in the agricultural sector than the other sectors. This analysis records evidence of Sectoral Rebound effects confirming Jevons’ paradox, whereby the economy’s energy mix which is strongly inclined towards non-renewable resources combined with the demand for green energy and efficiency, clearly supports the government’s recent policy moves in the direction of renewable energy generation and electricity substitution.

JEL Codes: O13, Q42, Q43, Q48

Keywords

Jevons’ paradox energy cost efficacy Granger causality Sectoral Rebound effects energy efficiency energy policy

1. Introduction

India’s electric power industry has recorded the major proportion of capital expenditure as a share of revenues. The central electricity authority has reported that 69.25% of the aggregate economy’s electricity generation capacity is thermal power by the end of March 2018. However, this generation is heavily dependent on coal inputs. The drop in the cost of installation of solar power plants in 2018 has raised the potential for widespread electric power generation yet, the low-capacity utilisation has raised recent concerns over the need for greater modernisation. The compound annual growth rate (CAGR) of consumption of all energy resources have been higher than the CAGR of production (from 2008–2009 till 2017–2018). The excess consumption has been satisfied by energy imports. The CAGR 2008–2009 to 2017–2018 of net imports¹ of coal stands at 13.68% and of crude oil at 5.20% (MOSPI, 2019). Figure 1 illustrates the economy’s energy mix and the pattern of consumption across renewable and non-renewable resources of energy. India continues to fulfill the mandate of the Paris Agreement on its renewable energy capacity. However, solar, wind and hydro power production remains substantially low (MOSPI, 2019). Then, where is India’s energy policy heading to? Is India revisiting Britain’s Coal question² via its experience of the Rebound effect across its sectors?

Figure 1.

Source: MOSPI (2019).

The energy mix indicates an economy wide dominance of consumption of non-renewable resources of energy while renewable energy contributes to only 13% of the total consumption. Though the production and availability of coal have recorded the highest growth among all commercial sources of energy, the fear of energy blackouts remains under concern for sustainable energy consumption.

A technological upgradation results in enhancement of resource use efficiency. Yet, the natural law governing consumption overpowers the productivity benefits of technical progress to ultimately raise the overall rate of that resource consumption. Stanley Jevons’ paradox postulates that the consumption of a resource increases when the efficiency of usage of that resource increases. Efficiency could be tied to technology-induced productivity shocks or government policies and schemes (Jevons, 1866). Thereby, low-cost energy efficient technologies drive up the Rebound effect in energy consumption. Investigating the presence of Jevons’ effect throws light on the sustainability implications of policies that target the energy sector. While Khazzoom–Brookes postulate in particular that micro-level efficiency improvements would induce macro level consumption of that energy resource, a Sectoral Rebound effect draws significance established from Jevons’ paradox that deals with this relationship primarily at the macro level (Khazzoom, 1980).

The rate of consumption of a resource increases with a greater consumer demand for it, while it can get substituted for other resources, on account of its alternative uses. A producer benefits from efficient technologies through lower cost of production which provides benefits of economies of scale. This induces large scale demand, thereby paradoxically leading to an increased consumption of the energy resource. While these are the direct spill-over benefits of efficiency improvement, the indirect benefits are in the form of diffusion of the technology across other industries and boosting aggregate consumption further.

Following the neo-classical theory, depletion of resources that pressurises energy prices upward also stimulates investments in technological developments. However, the paradox fails to materialise the benefits of energy efficiency as Jevons had pointed out at the steam engines. Such efficiency upgradation pushes for a fresh cycle of consumption wherein production units avail the benefits of economies of scale through input substitution as labour saving technologies raise capital intensity. The irreversibility of energy consumption holds the assessment of energy policies significant at different stages of economic development.

With rapid industrialisation, categorised as a semi-peripheral nation aiming at energy self-sufficiency, both economic and demographic expansions have contributed to a high material extraction rate in India since 1980. The impact of population explosion and economic growth on dematerialisation have been significant. However, the per capita extraction has remained below the world average. With an evident revelation of Jevons’ paradox, India and China have recorded high rates of extraction of natural resources (York et al., 2011).

British prosperity has largely been attributed to the Industrial revolution of the eighteenth century. However, this prosperity resulted from the availability of cheap and quality coal from Wales. Coal Question anchors this idea to energy efficiency and economic prosperity. Despite the scarce availability of coal, the consumption rate of coal was strikingly high. The consumption rate of coal was exponential, and with the invention of the more efficient steam engine, coal consumption should have reduced while the observations have pointed otherwise (Jevons, 1866).

The debottlenecking process is crucial to ensure the sustainability of limited resources, especially fossil fuels. Besides coal, extending the connotation of the paradox to other energy resources, adhering to the second law of thermodynamics, consumption of resources lead to environmental pollution and depletion. Efficiency policies assume the form of a loop that reinforces itself due to factors including rising pollution intensity due to increased consumption of natural resources and scarcity of resources which in itself induce more efficient measures to be adopted (Arrobbio & Padovan, 2018).

Adoption of 15 W energy efficient compact fluorescent lamps in the North Dakota region instead of 60 W incandescent light bulb reduced the energy needed, but on the contrary street lights were used for longer evening hours due to cheaper availability of electricity, thereby raising the overall demand (Bandyopadhyay, 2015). The insulation of household policy of Department of Energy and Climate Change by the UK government serves as an example of Direct Rebound effect, accounting for about 15% reduction in energy saving (Freire-González & Puig-Ventosa, 2015).

The distinct impacts of fuel efficiency on the consumption behaviour of different income classes, have been assessed in the US markets. The Energy Information Administration had reported that energy savings that resulted from fuel efficiency were temporary as it were being offset by the rate of vehicle use and ownership. Income effect has been stronger than the effect of gasoline price fluctuations in the personal transportation sector due to the relatively lower share of fuel prices in the overall consumer budget. The low fuel price elasticity of demand for private travel has ruled out the use of price channel to alter fuel consumption behaviour (Munyon et al., 2018).

The discrepancy between the expected and the actual outcomes of energy savings projects is a result of overlooking the behavioural notions of consumption that lead to ‘rebound effects’, thereby overestimating the benefits of efficiency enhancement in the form of energy consumption. This is evident from the efficient lighting projects of the Clean Development Mechanism in India (Gómez-Paredes et al., 2013).

As micro household characteristics of age, household size and employment status of the individuals significantly affect the consequences of vehicle fuel efficiency measured in terms of miles travelled, to what extent can macro sectoral effects of energy intensity and other determinants gain significance in drafting energy policies? While Polimeni and Polimeni (2006) suggest that the degree of Rebound effect varies across sectors and countries, would sector-wise energy efficiency gains degenerate environmental resources?

York has attempted to decouple economic growth from environmental damage by criticising the notion that ecological modernisation is a source of Green gross domestic product. His analysis has provided global evidence of Rebound effects where reduction of CO₂ intensity has paradoxically raised the level of emissions (York, 2010). Technological progress and industrial efficiency both have a significant impact on the mitigation of CO₂ emissions. However, Organisation for Economic Cooperation and Development countries experiencing excessive technological progress levels, also experience a CO₂ rebound. The story of emerging economies reads that rapid growth is accompanied by a substantial rebound (Wang et al., 2014; Wang & Wie, 2020).

Estimating Partial Rebound effects both in the long and short run, in urban China for electricity consumption, has indicated that the response of price elasticity of energy consumption was greater for a decline in prices compared to that of an increase. This result has been attributed to rapid economic growth in China due to a substantial rise in urban incomes and cheaper electricity availability (Wang et al., 2014).

A new type of urbanisation, that is, intensive production while ensuring ecological protection contrary to building huge infrastructure and subsequent population explosion, is becoming prominent in China. A spatial econometric analysis, conducted to measure its environmental impact has identified two diverse paths of impact. One, direct, by adopting new lifestyle changes such as complete household electrification and normalisations of Information Technologies, CO₂ emissions increase due to the need for more production. Second, indirect, by using energy-saving technologies, households have their financial pressure reduced, thereby using those technologies for longer hours and in larger quantities. This type of urbanisation has a paradoxical impact on CO₂ emissions and the said spatial distribution exhibits clustering in High–High and Low–Low patterns (Wang et al., 2019).

However, could emission rates be the only measure of technical efficiency in production when sectors adopt the output maximisation principle? Can a measure of the level of economic activity, as a proxy for technological improvements (efficiency in production) have an impact on the rate of consumption?

Cross country analyses have highlighted Jevons’ effect across geographically dispersed economies, provided the variations in the limited endowments of non-renewable resources. Asian economies experience Jevons’ paradox on the grounds of their structural transformation. Moreover, they have exhibited a relatively stronger Jevons’ effect when controlled for population (Polimeni & Polimeni, 2006). York and McGee (2015) have revealed that efficient economies have a significantly higher use of energy with conclusions regarding the need for higher energy to support the infrastructure, population and produce energy-efficient technologies. Empirical results establish the possible existence of Jevons’ paradox in India and Bangladesh with significant Backfire Rebound effects in non- renewable energy and coal consumption (Murshed, 2018).

Industrial energy conservation requires adequate study to understand the nature of the problem of energy consumption and the techniques required to preserve the ecological balance, provided its resilience.

Reducing the carbon intensity of gross domestic product (GDP) has been preferred to be cheaper than shifting to cleaner technologies of supply in India. Gupta and Sengupta (2013) have suggested demand-oriented policy instruments for energy savings with focus on the industrial sector’s response to changes in energy prices and technologies that permit inter-fuel substitution. A disaggregated analysis proved significant substitution possibilities between labour and energy and no complementarity between capital and energy. All the seven energy intensive industries have exhibited strong negative own price elasticity of demand for energy. Decomposition of total factor productivity is significantly energy price induced. Hence, tax on individual fuels is preferred over capital subsidies for energy conservation in India, due to insignificant capital price elasticity of demand for energy (Gupta & Sengupta, 2013).

Decomposition of sectoral energy intensity has highlighted low energy efficiency in the agricultural sector, though the aggregate energy intensity has been negative, from 1980 to 1996 (Bhattacharya & Paul, 2001).

Optimisation of industry specific factors has been found to have a significant impact on energy intensity. The positive effects of labour intensity, plant and machinery intensity and the negative effect of technology development intensity have been examined for different groups of industries categorised based on energy intensity (Soni et al., 2017).

Results of inter-state analysis have shown that states with a higher share of manufacturing output in energy-intensive industries have higher energy intensity and labour productivity has a positive relationship with energy efficiency. Power deficit has reduced energy intensity while power reforms have remained insignificant in regulating energy efficiency (Mukherjee, 2008).

The shift from coal traction to diesel and electric tractions in railways and increase in demand for road transport influenced energy consumption patterns (dominant usage of petroleum products) in the Indian transport sector. Energy consumption doubled in 1990 due to fuel efficiency improvements by 30% (from coal based thermal plants) (Ramanathan, 2007).

Transport energy consumption has been decomposed between freight and passenger transportation into energy intensity, transport volume and structural effects (modal shift and private transport demand). The negative intensity effect is stronger than the positive volume and structural effects (Tiwari & Gulati, 2013).

Sinha and Mishra (2019) used the first and second laws of thermodynamics to estimate energy efficiencies in the four sub-sectors of the transportation sector in India. With the sector’s consumption trend being strongly driven by roadways and railways, the overall energy efficiency along with the improvement potential factor have been increasing. Better energy utilisation can be enforced through improved public transportation and prioritising renewable energy sources (as the sector has recorded 78% energy losses).

Then, would India’s broad sectors, when treated as aggregate economic production units also face Jevons’ effect especially since the structural transformation in 1991?

With a structural shift in the agricultural output composition (output diversification from food crops to commercial crops), a rise in the degree of modernisation of farm inputs has altered the technical energy–output relationship in the Indian agricultural sector. Energy intensity and input–output mix differ across different crops and industrial goods. The production of different crops leaves environmental footprints of varied degrees (Polimeni & Polimeni, 2006).

While Green Revolution failed to satisfy the food demands of the population, which the Malthusian theory upheld, current debates about a second Green Revolution have raised questions about the duration of the impact of productivity shocks on the agricultural sector. Input–output analyses and the application of Engel’s curve³ to the Demand Bias hypothesis in justifying tertiarisation (service-led growth) lay emphasis on the impact of structural changes on energy intensity in the manufacturing sector (Ok et al., 2014; Tandon & Shahid, 2016). Severe air pollution holds private transport demands and vehicular pollution responsible for the worsening of Air Quality Index (Badami, 2006). Failing to comply with its commitment to ‘Carbon Sink’ (Sethi, 2018), India’s persistent consumption of coal, lignite and crude oil imposes serious risks about the optimal allocation of energy across its multitude of purposes as well as sectors. For a predominantly demand driven economy as India, the above three concerns question the sustainability of consumption of energy across India’s agriculture, manufacturing and transport sectors.

Can an evidence of Jevons’ paradox suggest focus on renewable energy resource supply rather than the traditional efficiency improvement approach, as the benefits of efficiency are questioned in the context of a rising deficit⁴ in the economy’s energy balance? With the economy’s potential to establish an equilibrium, due to persistent shocks arising due to rapid non-renewable energy depletion, exogenous injections are necessary catalysts to maintain energy sustainability. To hedge the risks of environmental damage from increased extraction of non-renewable energy resources, the economy has to draft policies that suggest alternatives for anchoring economic development.

II. Objective

The Energy Statistics, 2019 report stated that a decline in energy intensity for the aggregate economy has been attributed to faster growth in output rather than energy demand. This association arises from the fact that the growing share of services sector has induced electricity consumption due to a relatively faster growth in the former than the latter (MOSPI, 2019). However, an aggregate measure of energy intensity overlooks the sector-wise experiences of the technical efficiency of energy inputs. While literature has recorded evidences of Jevons’ paradox across energy sources for the aggregate economy, can Sectoral Rebound effects have a significant impact on the economy’s energy policy?

Amidst growing pressures on incorporating the circular economy model for energy production and consumption, the objective of this study is to empirically investigate Jevons’ paradox, individually across the three broad sectors of the Indian economy, namely, agriculture, manufacturing and transport sectors. The models aim at measuring the relative impact of sector-specific determinants of energy consumption and the magnitude of Rebound effects. This comparison across the sectors would facilitate drafting of appropriate energy policies pertinent to the sectors. These three models capture the temporal dimension of energy consumption rather than a spatial dimension due to unavailability of state-wise data on energy intensity and other variables of the three sectors.

III. Methodology and Data Variables

Three sector-specific models have been constructed, which have been estimated using an OLS technique using EViews software. The three broad sectors that have been chosen for the analysis include, agriculture, manufacturing and transport, as per the classification followed in the Energy Balance, presented in Energy Statistics report by the Ministry of Statistics and Programme Implementation (MOSPI). Energy consumption by these three sectors comprises of 80% of total energy consumption of the economy (excluding household consumption), as energy is treated as a production input. The time period chosen for the analysis covers 1984–1985 till 2016–2017. The variables filtered out from literature as determinants of energy consumption are energy intensity (000 MT per gross value added (GVA)), capital intensity (capital stock per employee), real GVA (at constant 2011–2012 prices), GDP per capita (at constant 2010 US $), crop yield (per hectare of arable land) and machinery intensity (tractors and power tillers per hectare of arable land).

Sector-wise energy consumption has been computed as an aggregate of all major commercial energy resources including coal, lignite, oil and electricity. Consumption of all these resources have been converted into one standard unit of measurement, petajoules. The maximum use of natural gas being in the fertiliser industry, it has not been included in the computation of energy consumption as fertilisers have not been accounted for in the measurement of energy intensity in the agricultural sector. Energy intensity is a proxy for energy use efficiency. A decrease in energy intensity is an indication of improvement in energy efficiency. This analysis attempts to utilise the newly published KLEMS Master Database (July 2019 update) (Erumban et al., 2019). Energy intensity, capital intensity and real GVA have been sourced from KLEMS, GDP per capita and machinery intensity from World Bank and crop yield has been computed by aggregating all food and commercial crop production per hectare of arable land. Data on sectoral consumption of various energy resources have been sourced from the Ministry of Petroleum and Natural Gas and the MOSPI.

IV. Empirical Analysis

This study aims at constructing a neo-classical model for a time-series investigation of Jevons’ effect across three broad sectors, as per the broad classification in Energy Statistics, published by the MOSPI. The model assumes a Cobb–Douglas production function as per the mathematical representation,

Q = f (K, L, E) = A \cdot K^{α} L^{β} E^{γ}

where, Q is the total output, K denotes capital, L denotes labour, E denotes energy, A is the constant, α, β and γ denote the share of capital, labour and energy, respectively. This model adopts the KLE⁵ production framework (Goldar et al., 2018). The model assumes technological progress as exogenous. The empirical analysis has been carried out sector-wise so as to capture the impact of sector-specific determinants of energy consumption besides the aggregate macroeconomic variables.

Unit Root Test Results

The variables have been tested for unit root using the Augmented Dickey–Fuller test and the results have been presented in Table 1. All the variables are integrated of the Order 1 and are hence I(1) processes except government expenditure in the transport sector, which is stationary at second difference. Therefore, the variables have been converted into their growth rates for estimation, which are stationary at level.

Table 1.

Unit Root Test Results.

Sector Variables	Agriculture
Sector Variables	Energy Consumption	Energy Intensity	Crop Yield	Machinery Intensity	Government Expenditure	Real GVA
At level	–3.17242	–2.334595	–2.624904	–1.677566	4.689481	–1.219802
At first difference	–7.939493***	–7.822066***	–5.489976***	–4.89061***	–7.518131***	–8.903259***
Growth rate	–6.904544***	–7.297512***	–5.424737***	–4.042485***	–6.090977***	–9.218926***
Variable	Manufacturing
t-Statistic	Energy Consumption	Energy Intensity	Capital Intensity	GDP Per Capita	Government Expenditure	Real GVA
At level	–0.230956	–0.981544	–0.522572	2.699655	0.770661	1.574853
At first difference	–5.106736***	–5.219416***	–3.024608***	–3.836594***	–4.963708***	–3.955134***
Growth rate	–5.149484***	–5.182869***	–2.744557*	–4.311333***	–8.475497***	–3.992289***
Variable	Transport
t-Statistic	Energy Consumption	Energy Intensity	Capital Intensity	Government Expenditure	Real GVA
At level	–2.160526	0.318094	–1.472764	7.20775	1.921984
At first difference	–6.155908***	–5.850135***	–2.699847*	–5.663622***^,a	–3.275334*
Growth rate	–4.722495***	–4.320312***	–2.630719*	–6.157833***	–3.042744**

Source: The authors.

Notes: GDP: Gross domestic product; GVA: Gross value added.

The table presents t-statistics for Augmented Dickey–Fuller tests.

***, ** and * denote significance at 1%, 5% and 10% levels, respectively.

^aAugmented Dickey–Fuller test at second difference.

Composition of Energy Consumption Across Sectors

The graphical illustrations presented in Figure 2 denote the composition of renewable and non-renewable energy resources in the three broad sectors. The transport sector and the manufacturing sector consume a greater proportion of non-renewable resources of coal, oil and lignite.

Figure 2.

Source: Computed by the authors from Indiastat.

Agricultural Sector

A fundamental graphical analysis is a pre-requisite to understand the trends of the variables of prime importance, energy intensity and energy consumption.

Both energy intensity and energy consumption have recorded a positive upward trend over time as shown in Figure 3. The cultivation of energy-intensive commercial crops has led to a rise in energy intensity in the agricultural sector. The use of commercial energy inputs per hectare of net sown area has increased during 1980–1981 till 2006–2007. The ratio of direct energy inputs remains dominant of the total energy inputs (Jha et al., 2012). Here, only direct energy use in agricultural sector has been taken into account (excluding pesticides and fertilisers).

Figure 3.

Source: KLEMS Master Database (RBI, 2019b).

Developing an empirical model to investigate Jevons’ paradox, the following equation has been estimated through OLS technique.

g r {[e c_{a}]}_{t} = α_{0} + β_{0} ​ g r {[e i_{a}]}_{t} + β_{1} g r {[c y]}_{t} + β_{2} g r {[g o v e x p]}_{t} + β_{3} g r {[g v a r]}_{t} + β_{4} g r {[m i]}_{t} + μ_{t}

(1)

The estimation results presented in Table 2 indicate that the rate of energy consumption in the agricultural sector is affected by energy intensity and the GVA by the sector. The coefficient of energy intensity facilitates the measurement of the degree of Rebound effect. Here, more than proportionate increase in energy consumption due to an increase in energy-use efficiency indicates that the Indian agricultural sector has been experiencing a Backfire Rebound effect wherein a 1% increase in energy intensity results in an increase in energy consumption by 1.499% (Figure 4). The crop yield, machinery intensity per hectare of arable land and government expenditure have an insignificant impact on energy consumption. Machinery intensity has been employed as a proxy for capital intensity in the agricultural sector. Though farm mechanisation, especially post the introduction of technological packages of Green Revolution in the Second Five Year plans, has influenced the energy-use pattern in Indian agriculture, the intensity of farm machinery (tractors and crushers) per hectare of land has no significant impact on the rate of energy consumption by this sector.

Table 2.

Estimation Results for Eq. (1).

Dependent Variable: Energy Consumption [gr (ec)]
Variables	β	t-Statistic
gr [ei]	1.499306***	3.295908 (0.454899)
gr [cy]	–0.946702	–1.497889 (0.632025)
gr [govexp]	0.002433	0.349136 (0.006970)
gr [gvar]	2.188689**	2.435779 (0.898558)
gr [mi]	0.126842	0.696675 (0.182067)
Constant (c)	–6.853084	–1.637070 (4.186188)
R ²	0.327163
Adjusted R²	0.197771
N	32
DW	1.789797

Source: The authors.

Notes: *** and ** denote significance at 1% and 5% levels, respectively.

Standard errors in parentheses.

cy: crop yield; ec: Energy consumption; ei: Energy intensity; govexp: Government expenditure; gvar: Real gross value added; mi: Machinery intensity.

Figure 4.

Source: The authors.

Jha et al. (2012) pointed out that agricultural output has a long run positive, cointegrating relationship with energy consumption, the real GVA has a statistically significant positive impact on energy consumption. Hence, demand for agricultural commodities raises demand for energy consumption in the sector, though crop yield resulting from a higher (increased) factor input productivity (land and labour) does not induce higher consumption of energy sources. The strength of output effect is stronger than the intensity effect.

Government capital expenditure in the agricultural sector has no significant influence on energy consumption. Literature fails to furnish any empirical evidence for the impact of government expenditure while there exist evidences to support correlations between public investments and capital intensity. The Indian agricultural sector suffers from a low research intensity due to lower investments on research and development in the sector. Hence, the demand for water saving technologies and efficient management of natural resources has been on the rise.

The investments (capital expenditures) directed at rural infrastructure (roads and rural electrification, credit and cooperatives) and irrigation projects have failed to contribute to capital formation due to the divergence of investments and expenditure. The rise in rural electrification has stimulated irrigation activities. Yet, the growth in the proportion of irrigated land has gained a slow momentum, thereby resulting in unutilised capacity and incomplete publicly funded projects. Further, electrification has only contributed to a rise in household electricity consumption whereas its share in the production process remains below its full potential. This has been the result of potential investments diverted towards unproductive revenue expenditure on subsidised agricultural inputs, which has crowded out private investments in the sector (Singh et al., 2015). The sector has faced low investment elasticity of agricultural GDP and the share of agricultural gross capital formation (GCF) to the total GCF has been declining, as shown in Table 3. Expenditures on centrally sponsored government schemes have primarily targeted social security, income support to farmers and employment generation, rather than energy consumption by tapping on farm technology which would raise output via the productivity channel, as shown in Table 4.

Table 3.

Share of Agricultural GCF to total GCF and Investment Elasticity of Agricultural GDP.

Year	Share of Agricultural GCF to Total GCF	Investment Elasticity of Agricultural GDP
6th Plan (1980–1985)	17.04	0.0538
7th Plan (1985–1989)	12.61	0.0584
8th Plan (1992–1997)	8.97	–0.0986
9th Plan (1997–2002)	9.66	1.0327
10th Plan (2002–2007)	8.02	0.2640

Source: Adapted from Marimuthu (2013).

Note: GCF: Gross capital formation.

Table 4.

Expenditure on Government Schemes.

Schemes	2017–2018 (in 000 Crore Rupees)	2018–2019 (in % to Total Expenditure)
Central Sector Schemes
PM-KISAN		0.81
Deen Dayal Upadhyaya Gram Jyoti Yojana	5	0.15
Ujjwala Yojana	2.3	0.13
Road works	36.8	1.66
Centrally sponsored schemes
MGNREGA	55.2	2.49
PM Krishi Sinchai Yojana	6.6	0.34
PM Gram Sadak Yojana	16.9	0.63

Source: Adapted from (RBI, 2019a).

Manufacturing Sector

Graphically, energy intensity has been declining while energy consumption has been increasing over time, in the manufacturing sector as shown in Figure 5.

Figure 5.

Source: Computed from the KLEMS Master Database and Indiastat.

The following model has been estimated to investigate Jevons’ paradox in the manufacturing sector.

g r : {[e c_{m}]}_{t} = α_{1} + λ_{0} g r {[e i_{m}]}_{t} + λ_{1} g r [{[k i n t]}_{t} + λ_{2} g r {[g o v e x p]}_{t} + λ_{3} g r {[g v a r]}_{t} + λ_{4} g r {[g d p p c]}_{t} + μ_{t}

(2)

The estimation results indicate that all the determinants are statistically significant (refer Table 5). The coefficient of energy intensity indicates that the manufacturing sector has been experiencing a Partial Rebound effect wherein a 1% increase in energy intensity leads to only 0.198% increase in energy consumption (Figure 6). This model captures the significant positive impact of government expenditure, real GVA and capital intensity. The sector has grown increasingly capital intensive, especially post the Liberalisation, Privatisation and Globalisation reforms of 1991 as the relative cost of inputs reduced (ceteris paribus), due to lower capital costs. The income effect (GDP per capita) is negative on the rate of energy consumption in the manufacturing sector. Income per capita serves as a structural parameter to examine the relationship between energy consumption and intensity in the manufacturing sector amidst the pre-mature de-industrialisation trend faced by the economy. This negative income effect is associated with rising incomes and tertiarisation which has shifted demand patterns away from manufacturing towards the services sector (Nayyar, 2012). The relative strength of input mix (capital intensity) is stronger than output effect which is in turn stronger than the intensity effect. Table 6 further clarifies that there exists no multicollinearity upon incorporating both the real GVA and the GDP per capita variables.

Table 5.

Estimation Results for Eq. (2).

Dependent Variable: Energy Consumption [gr (ec)]
Variables	λ		t-Statistic
gr [ei]	0.198548**		2.705249 (0.073394)
gr [kint]	0.395582**		2.078376 (0.190332)
gr [govexp]	0.023653***		2.829888 (0.008358)
gr [gvar]	0.274613*		1.834969 (0.149656)
gr [gdppc]	–0.601754*		–1.966304 (0.306033)
Constant (c)	4.157234***		3.018587 (1.377212)
R ²	0.465623	N	32
Adjusted R²	0.362858	DW	1.912049

Source: The authors.

Notes: ***, ** and * denote significance at 1%, 5% and 10% levels, respectively. Standard errors in parentheses.

gdppc: GDP per capita; kint: Capital intensity. Refer to notes in Table 2.

Table 6.

Variance Inflation Factors.

Variables	Co-efficient Variances	Centered VIFs
ei	0.005387	1.210064
kint	0.036226	1.163844
govexp	6.99E-05	1.031985
gvar	0.022397	1.731315
gdppc	0.093656	1.788691
c	1.896713	NA

Source: The authors.

Note: VIF: Variance inflation factor.

Figure 6.

Source: The authors.

Results indicate that higher government expenditures boost energy consumption in this sector. Government expenditure on social and physical infrastructures (road and rail connectivity) has stimulated a crowding of private investments in the industrial sector. This is a reiteration of the state’s role as highlighted in Hirschman’s theory of unbalanced growth with respect to investments made in high linkage sectors. The capital expenditure growth has been rising especially in the refineries, steel and cement industries. Government policies have been largely targeting the stimulation of economic activity in this sector through Make in India (MII) and the National Manufacturing Policy. The Perform-Achieve-Trade scheme post 2012 has been focusing on energy intensive sectors, which has ultimately led to an increased usage of energy resources in this sector (Goldar et al., 2018). The Ujjala and Uday schemes have predominantly been targeting electrical power efficiency in the industrial sector. A surge in investments following the Auto Fuel Policy and Vision 2025, released in 2003 with focus on fuel quality upgradation (shift to BS V automotive fuel), marked the significant growth in the automobile industry. Further, the relaxation of foreign direct investment (FDI) constraints has also raised the potential of the refineries sector to embark on an ‘Accelerated Transition Path’, contributing to the rise in energy efficiency as presented earlier in the graphical analysis(GOI, 2014).

Transport Sector

A graphical assessment of Figure 7 indicates that both energy intensity and energy consumption have recorded a downward trend.

Figure 7.

Source: Computed and generated by the authors from the KLEMS Master Database and Indiastat.

Empirically testing the paradox, the following model has been estimated through OLS technique.

g r {[e c_{t}]}_{t} = α_{2} + γ_{0} g r {[e i_{t}]}_{t} + γ 1 g r {[k i n t]}_{t} + γ 2 g r {[g o v e x p]}_{t} + γ 3 g r {[g v a r]}_{t} + μ_{t}

(3)

The empirical results indicate that all the determinants of energy consumption are statistically insignificant. The coefficient of energy intensity denotes that the transportation sector has been experiencing a Partial Rebound effect wherein a 1% increase in energy intensity leads to a 0.39% increase in energy consumption (Figure 8). The growing capital intensity of the sector has no impact on energy consumption as do the level of government expenditure and real output of the transport sector. India’s capital expenditure in the infrastructure sector is primarily directed towards railways and road transport systems (Table 7) (Sharma & Shah, 2018).

Table 7.

Capital Expenditure Heads (in ‘000 Crores).

Year	2016–2017	2017–2018	2018–2019
Major infrastructure	285	263	317
Energy	7	5	9
Indian Railways	45	43	53
Roads and bridges	44	54	64
Civil aviation	3	2	4
Irrigation	0.1	0.3	0.3

Source: Adapted from (RBI, 2019a).

Figure 8.

Source: The authors.

While capital expenditures target capacity expansion and modernisation of physical infrastructure, the pattern of revenue expenditure that is to follow the former also serves as an important determinant of energy consumption. Though capital expenditure raises the demand for energy sources, the proportion of revenue expended on fuel and electricity has remained relatively low while a major proportion is spent on maintenance and operation services (staff), indicating low operating efficiency. This sector has therefore been experiencing a low growth scenario wherein the proportion of capital expenditure is not in par with the operating expenditure (GOI, 2003).

Table 8 illustrates clearly the low share of fuel expenses to total working expenses with 17% of the revenue expenditure on fuel in 2016–2017 and 23% in 2018–2019 (Mishra, 2018). The upgradation from steam engines to diesel and electric locomotives has contributed to greater efficiency and hence the decline in energy intensity in this sector as presented graphically earlier. Since then, the railways had been accorded for an increase in transport output resulting from a rise in productivity levels. Yet, the major determinant of energy consumption remains energy intensity. With increasing energy efficiency, a continued capital expenditure on capacity creation to meet the growing demands of transport and connectivity by consumers and industries, can enable the sector to sustain energy consumption upon a shift to renewable energy. Both roadways and railways being heavily capital-driven, this sector requires either a planned management of diesel and electric tractions or expenditures in the form of alternative modes of traction using cleaner sources of energy. Only 12% of the capital expenditure was devoted towards electrification (Raghuram & Venkataraman, 2002).

Table 8.

Operating Fuel Expenses of the Indian Railways (in Crore Rupees).

Particulars	1996–1997	1997–1998	1998–1999	1999–2000	2000–2001	2001–2002
Operating expenses: fuel	3725.28	4637.45	4854.25	5629.67	6443.23	6910.94
TWE	21000.80	25876.03	27834.60	30843.99	34667.34	36470.44
% Share of fuel expenses to TWE	17.7387	17.9218	17.4396	18.252	18.5859	18.949

Source: Railway Budget 2002–2003, Ministry of Railways, Government of India.

Note: TWE: Total working expenses.

Only 1% of electricity consumption in the economy is by the transport sector (MOSPI, 2019). Capital expenditures in this sector are made ex-ante demand, rather than ex-post. Hence, the actual realisation of traffic determines the degree of capacity utilisation and the subsequent energy demands. Capital expenditures on gauge conversion and rail electrification have been reported as ‘unremunerative’ and hence, remains as underutilised capital (GOI, 2001). Additionally, the persistent dominance of oil as a major energy resource being consumed, leaves the sector vulnerable to fluctuations in energy prices in the world market.

Energy Cost-efficiency and Energy Intensity

Jevons highlighted the cost-efficient productivity benefits from an industry’s adaptation to technological upgradations. Hence, associating energy-use efficacy with cost efficacy, the following graphical illustrations and analysis are complementary to the regression results obtained from the previous model estimations (refer Figure 9).

Figure 9.

Source: Computed and generated by the authors from the KLEMS Master Database (RBI, 2019b).

Polimeni and Polimeni (2006) analysed that the real energy price elasticity of energy demand has a significant impact on energy consumption and affects farm profitability due to high diesel and electricity price indices. When an industry experiences technical progress, the cost of production reduces and it enjoys economies of scale. Goldar et al. (2018) studied the impact of a decline in real energy prices on energy consumption whose results highlighted the significant substitution effect which raised energy intensity. Similarly, this real energy prices effect has been measured using energy cost per unit of energy consumed (in crores per petajoules). This adds additional insights to the comparison of empirical results and the graphical trends of energy intensity and energy consumption across the three sectors.

Agricultural Sector

While the estimation of Equation (1) indicates a Backfire Rebound effect, Figure 3 does not significantly indicate the presence of Jevons’ effect. Complying with Jevons’ primary focus on the rate of energy consumption, Equation (1) has quantified the magnitude of Jevons’ effect, while a decline in the energy cost has led to a substitution of energy for capital. This has raised the energy intensity which has exerted an upward pressure on energy consumption. This is further validated by the insignificant impact of machinery intensity on energy consumption (Table 2).

Manufacturing Sector

As represented by the graph in Figure 5, an increase in both energy efficiency and energy consumption clearly signifies the presence of Jevons’ effect. However, the magnitude of the Partial Rebound effect is relatively lower than the other sectors. While an increase in the cost per unit of energy consumption ceteris paribus should have technically reduced energy consumption, the relatively dominant effects of capital intensity and output have stimulated a demand-pull effect on energy consumption.

Transport Sector

While an increase in energy efficiency and a decline in energy consumption indicate a reversal of Jevons’ effect, their growth rate trends have exhibited a Partial Rebound effect. The rise in cost per unit of energy consumption has significantly exerted a downward pressure on energy consumption, as the sector lacks any other strong demand forces (results from Table 9).

Table 9.

Estimation Results for Eq. (3).

Dependent Variable: Energy Consumption [gr (ec)]
Variables	γ		t-Statistic
gr [ei]	0.385302*		1.899871 (0.202804)
gr [kint]	–0.137069		–0.525743 (0.260715)
gr [govexp]	–0.007550		–1.252306 (0.006029)
gr [gvar]	0.026859		0.046836 (0.573483)
Constant (c)	5.270189		1.286684 (4.095946)
R ²	0.174378	Adjusted R²	0.052063
N	32	DW	1.769342

Source: The authors.

Notes: * denotes significance at 10% level. Standard errors in parentheses. Refer to notes in Tables 2 and 5.

Granger Causality

Murshed (2018) has highlighted the absence of long-run causal relationship between energy consumption and intensity variables of primary and secondary energy consumption, in India and Bangladesh, while recording bidirectional causality only for population and FDIs with energy consumption. Granger causality tests were run to investigate the presence of long-run causal association between energy consumption of different sectors and the sector-specific determinants of energy demand.

None of the sectors exhibit any significant causal association among the variables (Tables 10 11 12).

Table 10.

Granger Causality Test Results for the Agricultural Sector.

Null Hypothesis	F- Statistic
ei does not Granger Cause ec	1.0338
ec does not Granger Cause ei	2.6884*
mi does not Granger Cause ec	0.0409
ec does not Granger Cause mi	0.5832
gvar does not Granger Cause ec	0.1116
ec does not Granger Cause gvar	1.8393
cy does not Granger Cause ec	0.5010
ec does not Granger Cause cy	1.5190
govexp does not Granger Cause ec	0.5657
ec does not Granger Cause govexp	0.1400

Source: The authors.

Note: * denotes significance at 10% level. Refer to notes in Tables 2 and 5.

Table 11.

Granger Causality Test Results for the Manufacturing Sector.

Null Hypothesis	F-Statistic
ei does not Granger Cause ec	0.2625
ec does not Granger Cause ei	1.6487
govexp does not Granger Cause ec	0.1380
ec does not Granger Cause govexp	1.2306
gvar does not Granger Cause ec	0.7409
ec does not Granger Cause gvar	0.2017
kint does not Granger Cause ec	0.7033
ec does not Granger Cause kint	0.1295
gdppc does not Granger Cause ec	0.2929
ec does not Granger Cause gdppc	0.2766

Source: The authors.

Table 12.

Granger Causality Test Results for the Transport Sector.

Null Hypothesis	F-Statistic
ei does not Granger Cause ec	0.2702
ec does not Granger Cause ei	1.1029
kint does not Granger Cause ec	0.0267
ec does not Granger Cause kint	1.0280
govexp does not Granger Cause ec	2.1045
ec does not Granger Cause govexp	0.0037
gvar does not Granger Cause ec	1.4759
ec does not Granger Cause gvar	1.9409

Source: The authors.

V. Policy Implications

Both the agricultural and the manufacturing sectors have depicted a relatively stronger output effect, in confirmation with the results of literature, regarding the growth in output over that of energy demand post 2010 (Goldar et al., 2018). The manufacturing sector has witnessed the most prominent decline in energy intensity while continuing to remain the largest consumer of non-renewable fossil fuels (coal, lignite and oil). The degree of Rebound effect is higher in the transportation sector than the manufacturing sector. A Backfire Rebound effect experienced by the agricultural sector requires greater policy attention. The rising pressure on agricultural output in the context of the Malthusian theory, and the positive relationship between real output and energy consumption is to be brought under serious concern. Hence, the agricultural sector strongly requires the energy policy to focus on substituting non-renewable fossil fuel sources with renewable energy generation, to meet its demand for growing output and energy requirements. This would be relatively a more sustainable policy shift rather than efficiency enhancement for India to progress in achieving its target of becoming a power-surplus economy.

The portfolio of government expenditure remains inclined towards employment generation, industrial infrastructure and growth, while the expenditure on energy forms only 6% of the total capital outlay on infrastructure. Hence, the focus of the policies should rather be the rate of consumption and a shift in consumption behaviour, which insists on the need for government expenditure. While efficient utilisation of the existing non-renewable resources leads to a higher rate of consumption, a shift to green energy can meet the rising consumption demands while also preventing environmental degradation. This evidence acts as a supplementary support to the New Energy Policy’s objectives of substituting electricity for fossil fuel energy.

NITI Aayog’s draft New Energy Policy (2019) brought the budgetary proposal of electric vehicles under the ambit of MII, thereby continuing to focus on energy efficiency. The policy’s ultimate focus lays on energy transition via channelling domestic investments into electricity generation so as to divert demand away from non-renewable fossil fuel energy sources (Gupta, 2019; Aayog, 2017).

The One Nation, One Grid (2019) policy towards raising the economy’s self- sufficiency in the power sector by centralising power distribution and facilitating regional integration, eventually aims at meeting the rising energy consumption demand (GOI, 2019). Recent trends as reported by the Energy statistics describe the increased rate of consumption of fossil-fuel sources. India continues to witness a service-led growth under pre-mature de-industrialisation wherein the industrial sector remains the dominant consumer of energy, followed by domestic purposes. Consumption of coal and lignite remains the highest in India with the per capita consumption of energy following a constantly increasing trend (MOSPI, 2019).

VI. Limitations of the Study

Despite emphasis laid by Jevons and Polimeni (2008), population (demographic) effects could not be captured in the above models due to methodological constraints.

The low R² and adjusted R² estimates have yet been considered for the study, as this study focuses primarily on investigating the presence of Jevons’ effect and not on the estimation of a complete model. Econometric analysis at the firm level has shown that energy intensity is dependent on the size, age, labour, research intensity and various other disaggregated factors pertaining to a firm (Sahu & Narayanan, 2011). This study has only identified major aggregate economic determinants pertaining to the broad sectors as a fundamental empirical analysis of the paradox.

The post 2010 decline in real energy prices has led to a substitution of energy for capital in the contribution to output growth, thereby raising the level of energy intensity. While oil prices have had a significant impact on agricultural farm profitability in India, the models failed to measure the impact of energy prices on the rate of consumption across sectors. The role of the government in investments in the energy sector has only been captured using capital expenditure while it has failed to analyse the legal and institutional effects of tariffs, energy efficiency standards, laws and taxes (emission tax in the transportation sector).

VII. Conclusion

While the graphical illustrations reveal the heterogenous experiences of the three sectors, empirical testing of Jevons’ postulation has used the rate of growth of intensity and consumption rather than the absolute levels. As the neo-classical theory suggests, depletion of non-renewable resources raises its price that attracts investments, which under the assumption of the Embodiment hypothesis,⁶ increases the energy resource use-efficiency through a Vintage effect. This continues to raise the consumption of the resource leading to further depletion. Hence, this cycle has to be broken in order to divert resources towards the generation and supply of renewable sources of energy. As efficiency improvements in the consumption of non-renewable resources result in scale effects which diminish environmental resources, a shift to renewable energy enables sectors to avail the short-term benefits of substitution effects, while long run benefits accrue to the aggregate economy in the form of a transition to sustainability. In the light of the ‘exhaustive’ consequences of modernisation, ‘economic development’, which is contained in sectoral growth, should be conceptualised to enhance the quality and sustainability of production rather than the scale of production. Herein, the energy efficiency–energy consumption nexus emerges as the cornerstone of energy policy.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

Notes

ORCID iDs

Poovitha Murugasamy

Vandita Mishra

References

Aayog

(2017). Draft National Energy Policy . Government of India.

Arrobbio

, & Padovan

(2018). A vicious tenacity: The efficiency strategy confronted with the Rebound effect. Frontiers in Energy Research , 6, 114. https://doi.org/10.3389/fenrg.2018.00114

Badami

M. G.

(2005). Transport and urban air pollution in India. Environmental Management , 36(2), 195–204. https://doi.org/10.1007/s00267-004-0106-x

Bandyopadhyay

(2015). Careful with your energy efficiency program! It may ‘rebound’! Clean Tech Environmental Policy , 17(6), 1381–1382. https://doi.org/10.1007/s10098-015-1002-1

Bhattacharya

, & Paul

(2001). Sectoral changes in consumption and intensity of energy in India. Indian Economic Review , 36(2), 381–392. https://ideas.repec.org/a/dse/indecr/v36y2001i2p381-392.html

Erumban

A. A.

, Aggarwal

, Das

P. C.

, & Das

D. K.

(2019). Measuring productivity at the industry level – The India KLEMS Database . Reserve Bank of India. https://m.rbi.org.in/Scripts/BS_ViewPublicationReport.aspx

Freire-González

, & Puig-Ventosa

(2015). Energy efficiency policies and the Jevons Paradox. International Journal of Energy Economics and Policy , 5(1), 69–79. https://ideas.repec.org/a/eco/journ2/2015-01-06.html

GOI. (2001). The Indian Railways Report . Ministry of Railways. https://shodhganga.inflibnet.ac.in/bitstream/10603/52950/10/10_chapter%204.pdf

GOI. (2003). Working expenditures of Indian Railways . Ministry of Railways.

10.

GOI. (2014). Report of the Expert Committee on Auto Fuel Vision & Policy 2025 . http://petroleum.nic.in/sites/default/files/autopol.pdf

11.

GOI. (2019). Union Budget 2019-20: Delivering growth, forging ahead . Government of India.

12.

Goldar

, Aggarwal

, & Das

P. C.

(2018, June). Intensity and use of energy in Indian industries and the contribution of energy to growth. In Fifth World KLEMS Conference. https://scholar.harvard.edu/files/jorgenson/files/pl09b_intensity_and_use_of_energy_in_indian_industries.pdf?m=1528208475

13.

Gómez-Paredes

, Ishihara

K. N.

, Okumura

, & Yamasue

(2013). Energy efficiency to reduce poverty and emissions: A silver bullet or wishful thinking? Analysis of efficient lighting CDM projects in India. Procedia Environmental Sciences , 17, 547–556. https://doi.org/10.1016/j.proenv.2013.02.070

14.

Gupta

, & Sengupta

(2013). Energy savings potential and policy for energy conservation in selected Indian manufacturing industries. Review of Market Integration , 5(3), 363–388. https://doi.org/10.1177/0974929214538361

15.

Gupta

(2019, November 27). Government in the process of finalizing new energy policy: NITI Aayog Additional Secretary. pv magazine . https://www.pv-magazine-india.com/2019/11/27/government-in-the-process-of-finalizing-new-energy-policy-niti-aayog-additional-secretary/

16.

IEA. (n.d.). IEA Atlas of Energy – Country profiles . International Energy Agency. http://energyatlas.iea.org/#!/profile/WORLD/IND

17.

Intriligator

M. D.

(1992). Productivity and the embodiment of technical progress. The Scandinavian Journal of Economics , 94, S75–S87. http://www.jstor.com/stable/3440248

18.

Jevons

W. S.

(1866). The Coal Question: Concerning the progress of the nation, and the probable question of our Coal mines . Macmillan and Co. https://oll.libertyfund.org/titles/317

19.

Jha

K. G.

, Pal

, & Singh

(2012). Changing energy-use pattern and the demand projection for Indian agriculture. Agricultural Economics Research Review , 25(1), 61–68. https://doi.org/10.22004/ag.econ.126043

20.

Khazzoom

(1980). Economic implications of mandated efficiency in standards or household appliances. Energy Journal , 1(4), 21–40. www-jstor-org.web.bisu.edu.cn/stable/41321476

21.

Marimuthu

(2013). Gross capital formation and GDP growth in Indian agriculture sector. Munich Personal RePEc Archive , 9.

22.

Mishra

(2018). Demand for Grants 2018-19 Analysis: Railways . Institute for Policy Research Studies. https://www.prsindia.org/sites/default/files/budget_files/Railway%20Budget%20Analysis%202018-19.pdf

23.

MOSPI. (2019). Energy statistics . Ministry of Statistics and Programme Implementation, Central Statistics Office.

24.

Mukherjee

(2008). Energy use efficiency in the Indian manufacturing sector: An interstate Analysis. Energy Policy , 36, 662–672. https://doi.org/10.1016/j.enpol.2007.10.015

25.

Munyon

V. V.

, Bowen

, & Holcombe

(2018). Vehicle fuel economy and vehicle miles travelled: An empirical investigation of Jevons' Paradox. Energy Research & Social Sciences , 38, 19–27. https://doi.org/10.1016/j.erss.2018.01.007

26.

Murshed

(2018). Revisiting the Jevons Paradox of energy economics: Empirical evidence from Bangladesh and India. International Review of Business Research Papers , 14(1), 68–93. https://ssrn.com/abstract=3161649

27.

Nayyar

(2012). The demand for services in India. A mirror image of Engel's law for food? The Service Sector in India's Development , 65–91. https://doi.org/10.1017/CBO9781139095860.005

28.

, Lee

, & Kim

(2014). Service-led catch-up in the Indian economy: Alternative hypotheses on tertiarization and the leapfrogging thesis. Seoul Journal of Economics , 27(1), 1–40. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2423198

29.

Polimeni

(2008). Jevons' Paradox: The cases of India and China. International Journal of Interdisciplinary Social Sciences , 3, 145–157. https://www.researchgate.net/publication/286160984_Jevons'_paradox_The_cases_of_India_and_China

30.

Polimeni

, & Polimeni

R. I.

(2006). Jevons' Paradox and the myth of technological liberation. Ecological Complexity , 3(4), 344–353. https://doi.org/10.1016/j.ecocom.2007.02.008

31.

Raghuram

, & Venkataraman

(2002). Public expenditure accountability of the Indian railways. CiteSeerX . http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.579.6714&rep=rep1&type=pdf

32.

Ramanathan

(2007). Indian transport sector: Energy and environmental implications. Energy Sources , 18(7), 791–805. https://doi.org/10.1080/00908319608908811

33.

RBI. (2019a, September). Expenditure on major government schemes . Union Budget 2019-20: An Assessment. https://www.rbi.org.in/Scripts/BS_ViewBulletin.aspx?Id=18467

34.

RBI. (2019b, July). KLEMS Database . Reserve Bank of India. https://www.rbi.org.in/Scripts/KLEMS.aspx

35.

Sahu

S. K.

, & Narayanan

(2011). Determinants of energy intensity in Indian manufacturing industries: A firm level analysis. Eurasian Journal of Business and Economics , 4(8), 13–30. https://www.ejbe.org/EJBE2011Vol04No08p013SAHU-NARAYANAN.pdf

36.

Sethi

(2018, November 12). India set to achieve 2 of 3 Paris Agreement Goals: Draft Government Report. The Wire . https://thewire.in/environment/india-paris-climate-agreement-targets

37.

Sharma

, & Shah

(2018, March). Industrial Capex: On recovery path. Motilal Oswal . https://www.motilaloswal.com/site/rreports/HTML/636576524030283986/index.htm

38.

Singh

, Pal

, & Jha

G. K.

(2015). Transitioning India's public expenditure in agriculture towards higher growth and equity. Indian Journal of Agricultural Economics , 70(3), 1–13. https://doi.org/10.22004/ag.econ.230050

39.

Sinha

, & Mishra

(2019). Energy consumption pattern of Indian transportation sector and its thermodynamic analysis. Journal of Scientific and Industrial Research , 78, 520–525. http://nopr.niscair.res.in/bitstream/123456789/49558/1/JSIR%2078%288%29%20520-525.pdf

40.

Soni

, Mittal

, & Kapshe

(2017). Energy intensity analysis of Indian manufacturing industries. Resource-Efficient Technologies , 3(3), 353–357. https://doi.org/10.1016/j.reffit.2017.04.009

41.

Tandon

, & Shahid

(2016). The changing energy intensity in Indian economy: A sector-level analysis based on Input–Output Model. International Journal of Energy Economics and Policy , 6(3), 449–462. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2817747

42.

Tiwari

, & Gulati

(2013). An analysis of trends in passenger and freight transport energy consumption in India. Research in Transportation Economics , 38, 84–90. https://doi.org/10.1016/j.retrec.2012.05.003

43.

Wang

, & Wei

(2020). Coordinating technological progress and environmental regulation in CO₂ mitigation: The optimal levels for OECD countries & emerging economies. Energy Economics , 87, 104510. https://doi.org/10.1016/j.eneco.2019.104510

44.

Wang

, Lu

, & Wang

J.-C.

(2014). Direct rebound effect on urban residential electricity use: An empirical study in China. Renewable and Sustainable Energy Reviews , 30(C), 124–132. https://doi.org/10.1016/j.rser.2013.09.002

45.

Wang

, Sun

, & Wang

(2019). How does the new type of urbanization affect CO₂ emission in China? An empirical analysis from the perspective of technological progress. Energy Economics , 80, 917–927. https://doi.org/10.1016/j.eneco.2019.02.017

46.

York

(2010). The Paradox at the heart of modernity. International Journal of Sociology , 40(2), 6–22. www-jstor-org.web.bisu.edu.cn/stable/20788543

47.

York

, Ergas

, Rosa

, & Dietz

(2011). It's a material world: Trends in material extraction in China, India, Indonesia and Japan. Nature and Culture , 6(2), 103–122. https://doi.org/10.3167/nc.2011.060201

48.

York

, & McGee

J. A.

(2015). Understanding the Jevons Paradox. Environmental Sociology , 2, 1–11. https://doi.org/10.1080/23251042.2015.1106060