Complementarity between labour and energy in Indian manufacturing

As India develops its carbon markets to mitigate climate change, it is important to consider the short-term costs for industry. Analysing data on formal manufacturing from 2009-10 to 2019-20, this article finds that energy and labour are complementary inputs in production – if an intervention such as a carbon tax raises the cost of energy and reduces its use in manufacturing, employment will go down as well.

A carbon tax on fossil energy has been proposed as an efficient instrument to mitigate climate change. However, in an emerging economy such as India, a carbon tax-induced increase in prices for fossil energy could generate economic costs in the short run. Policymakers have raised concerns that a carbon tax could hurt industrial competitiveness and reduce employment.

In recent research (Aggarwal 2024), I analyse the effects of increases in prices for fossil-based energy (coal, and electricity derived from coal) on the ratio of labour-to-energy use in firms in the manufacturing sector. If the ratio of labour-to-energy use (the ‘elasticity of substitution’) increases by less than one in response to a 1% increase in the price of coal or electricity (relative to the workers’ average wage), it suggests that labour and energy may be complements in industrial production. If the elasticity of substitution between labour and energy is less than one, it indicates that a carbon tax imposed on coal or electricity could reduce employment in the manufacturing sector in the short-run.¹Whether the short-run adverse effects of a carbon tax can be ameliorated through government intervention is a relevant concern for domestic climate policy.

India’s coal market

To measure the elasticity of substitution between labour and energy, I utilise external variation in coal and electricity prices over the 2009-10 to 2019-20 period. I combine annual data on energy prices from the annual Coal Directories and Ministry of Power with firm-level panel data from the Annual Survey of Industries (ASI) for the corresponding period. The analysis relies on three sources of geographical variation in energy prices. First, the mining cost of coal is drawn from “pithead” price set by Coal India Limited (CIL) for coking and non-coking coal. The government-owned CIL is India’s largest coal producer, mining over 80% of raw coal in the country, and remains a profitable company (Tongia et al. 2020). While coking coal is used directly in the blast furnace for steel production, non-coking coal is used for captive power generation by firms and in several other sectors (Ministry of Coal, 2021).

The second source of variation in coal prices arises from freight charges for coal transportation by the Indian Railways. The Railways annually revise the freight charges for coal transportation by distance travelled (for example, 0-50 kms, 51-100 kms… up to 3,000-3,500 kms). The distances between each of approximately 50 mines and the district of highest coal purchases in each state, from the ASI firm survey, are calculated. This generates approximately 1,500 pair-wise distances. Then, freight charges are calculated based on the shortest distance to the mine, for each state. Since the majority of coal mines are located in Eastern India (with a few mines operating in Maharashtra, Gujarat and other states), we observe that coal is transported over 1,400 kms on average from mines to factories. Figure 1 displays the geographical distribution of current coal reserves (Central Mine Planning & Design Institute, 2023). The average annual prices of coking and non-coking coal (the sum of the pithead price and the freight charges) are displayed in Figure 2.

Figure 1. Inventory of coal reserves in India, 2023

Source: Central Mine Planning & Design Institute (2023).

Finally, I utilise variation in India’s introduction of the “Clean Energy Cess” on coal, which commenced at Rs. 50/tonne of coal in 2010, and has gradually been raised to Rs. 400/tonne of coal in 2016. The price of coal is thus constructed at the state-level as the sum of the (i) mining cost, (ii) freight charge and (iii) coal cess. For electricity prices, I construct ‘a shift-share instrumental variable’ (SSIV), which leverages variation in annual increases in coal prices set by CIL, with the pass-through of coal prices to electricity prices determined by the share of coal in power generation for each state. The SSIV is the average annual price of coal at the state level, interacted with the “initial share of coal in electricity generation”, for each state (Abeberese 2017, Singer 2024). The initial shares refer to the 2009-10 period, prior to the introduction of the national coal cess. Thus, the SSIV provides a source of exogeneous variation in electricity prices induced by increases in state-level coal prices.

Figure 2. Average annual prices of coking and non-coking coal, 2010-2020

Note: The average prices displayed are calculated as the sum of the pithead price (notified by CIL) and the freight charge (notified by the Indian Railways) for the shortest distance between mines and states.

Labour and energy are complements

Across the manufacturing sector, I find labour and energy (coal, and electricity derived from coal) are complements in production. The elasticity of substitution is precisely estimated to be 0.3-0.4 across the two fuels (for more details, see Tables 3 and 4 in Aggarwal (2024)). Similarly, we find labour and energy are complements in all NIC two-digit manufacturing sectors², except the tobacco industry. Further analysis of heterogeneous effects shows elasticities across the firm size distribution (measured as firm-level output and employment) are positive and less than one, which suggests that labour and energy are complements in the manufacturing sector.³

Given the estimated elasticity of substitution is below one in all manufacturing sectors in the formal, organised industry, it raises concerns of potential reductions in employment in response to a carbon tax on coal and electricity in India. Policy discussions ought to focus on mechanisms to compensate firms, particularly small- and medium-sized firms, for the short-run costs of carbon taxes (whether implemented as a carbon price or as an emissions trading scheme) in the Indian context. This is particularly relevant given India’s developing carbon market and the launch of the Carbon Credit Trading Scheme, by the Bureau of Energy Efficiency, Government of India.

The shift from coal to renewable energy in power generation

Over the last decade or so, the share of renewable energy in installed capacity for electricity generation has steadily risen and stands at over 40% (Ministry of Power, 2022). A pertinent question is whether we might observe substitution from coal towards renewable energy for power generation in the context of a domestic carbon tax. To analyse this question, I estimate the elasticity of substitution between coal and renewable energy in power generation.⁴Drawing on annual data on the installed generation capacity by energy source at the state-level for the 2009-10 to 2019-20 period, I estimate the degree of substitution between coal and renewables (σ_CR) (For the full results, see Table 8 in Aggarwal (2024)).

As expected, the elasticity of substitution between coal and renewables is greater than one (σ_CR= 1.39). This suggests a carbon tax on coal, which raises the cost of fossil energy relative to the cost of renewable energy for power generation, is likely to accelerate the shift towards renewable energy and facilitate decarbonisation of the electric grid.

Reconciling the two results implies a need for targeted policy interventions to compensate the manufacturing sector for potential losses in employment due to a carbon tax, while strategically deploying carbon taxes in the power sector to facilitate the low-carbon transition in the power sector and the overall economy.

Notes:

This is the case if higher output levels at a firm are associated with increased use of coal or electricity in production.
Sectors are defined according to National Industrial Classification (NIC) codes. NIC codes are a statistical standard for developing and maintaining a comparable data base for various economic activities, developed with an intent to ascertain and analyse as to how each economic activity is contributing towards national income.
A similar study (Bretschger and Jo 2024) finds labour and energy are complements in French manufacturing firms.
I follow the approach by Papageorgiou et al. (2017), who estimate the elasticity of substitution between clean and dirty energy for 26 countries.