by Ajay Shah and Susan Thomas.
The problem
The Indian electricity system has major problems. It suffers from a central planning problem, where officials control the resource allocation, which undermines efficiency and innovation. It has a carbon emissions problem, with the lack of an effective path into the clean energy transition. It has a public finance problem, where debt sustainability in some states is materially affected by theft and by subsidies are paid for by the exchequer. As an example, in Mehta et al. 2024, we show that for Tamil Nadu, "A complete electricity sector reform versus business-as-usual translates into an FY 2028 outcome for the debt/GSDP ratio of 32.47% vs. 43.53%, and an IP/RR ratio outcome of 19.71% vs. 26.12%".
A root cause of these difficulties is subsidised and stolen electricity. As there is no free lunch, the lost revenues have to show up as a combination of explicit budgetary allocations for electricity subsidies, or distress for the distribution company. The precise mechanisms through which electricity is stolen are surprisingly subtle, e.g. as shown in Mahadevan 2024, which casts a shadow on conventional measures of AT&C losses. While there is some movement in favour of more transparent on-budget subsidies (Jaitly and Shah 2024) in states such as Karnataka, the overall problem reflects a combination of transparent on-budget subsidies, weak bill collection, and theft.
 |
| Figure 1: Feedback loops in the Indian electricity system Source: Figure 10 (page 15) from Jaitly et al. 2025. |
As Figure 1 shows, there are multiple positive feedback loops in operation in the Indian electricity system, which are grinding away, worsening the distress. Inadequate payment leads to SEB distress, which in turn forces high tariffs on paying C&I consumers, driving them to exit the grid and further worsening SEB finances. Inadequate payment for electricity, by many firms and households, is the core problem which then plays out in various ways.
The developments in Pakistan in recent years give us illustrations of how such causal forces could play out in the future in India (Economic Times, 2025). Conversely, the imposition of a single price for electricity applied to all buyers of electricity would materially change these feedback loops by alleviating financial distress in electricity distribution.
Household data as a research tool
A significant amount of theft of electricity is surely done by firms, who have high incentive to put in effort to obtain stolen electricity. But the overt political problems of subsidised electricity for households or agriculturists, and the political economy problems of a large number of persons stealing electricity, are uniquely present in the household sector. Household survey data offers valuable knowledge about the problems. Instead of starting from budget disclosures and the data as reported by distribution companies, we go bottom up by asking households what they pay for electricity. There are grounds for trusting the CMIE CPHS measurement of electricity expenditures at the household level (Das et al. 2024).
Figure 11 (page 19) from Jaitly et al. 2025 shows how electricity expenditures in the household data in Tamil Nadu are unusually low by Indian standards. While this appears out of line when we think that Tamil Nadu is richer than the overall Indian average, there is a need for more careful analysis which compares similar households and juxtaposes different arrangements for electricity distribution.
The gains from private distribution
In the present research, we focus on two groups of the biggest Indian cities with alternative electricity distribution arrangements. Bombay, Delhi and Calcutta have private distribution. Bangalore and Madras have public sector distribution. Using the CMIE CPHS data, we work out the median household expenditure on electricity in these two groups of cities. These are large datasets: In 2024, Bangalore and Madras add up to 1,641 households and Bombay, Calcutta and Delhi add up to 2,667 households. Given these large sample sizes, the median estimates are statistically robust. Weighted estimates are used, which can be interpreted as populated-weighting across the cities.
 |
| Figure 2: Median household electricity expenditure in large cities, private vs. public electricity distribution Source: Authors' calculations using CMIE CPHS data |
While we may broadly think that household prosperity in Bombay, Delhi and Calcutta is similar to that seen in Bangalore and Madras, we control for this by placing households into nationwide urban consumption quartiles.
| Quartile | Median Cons. | Median electricity (private) | Median electricity (public) |
|---|---|---|---|
| (monthly) | (monthly spend) | (monthly spend) | |
| Q1 (poor) | 10,536 | 500 | 0 |
| Q2 | 14,755 | 665 | 0 |
| Q3 | 19,307 | 775 | 0 |
| Q4 (rich) | 29,522 | 1500 | 75 |
The quartiles are formed based on the all-India distribution of urban household consumption as seen in the CMIE CPHS data in 2024. The all-India median urban consumption runs from Rs.10,536 a month for the poorest quartile to Rs.29,522 a month for the richest quartile. With this in hand, all the urban households in Bombay, Delhi, Calcutta and then Bangalore and Madras are placed into the appropriate quartile bins. Since the bins are based on all-India urban consumption, households in each group (private distribution/public distribution) will not be equally split across the quartile bins. As we are studying the biggest cities in India, more households are likely to be slotted in higher consumption bins. We report the median value of the monthly electricity expenditure for the two groups.
These results show that after controlling for affluence of the households, public sector distribution obtains much lower payments for electricity when compared with private sector distribution. In the future, this work needs to be made statistically more rigorous by setting up a matching scheme where households in Bangalore/Madras are matched to households in Bombay/Delhi/Calcutta based on asset ownership. However, the magnitude of the difference suggests the core finding is robust.
The political economy of the Indian electricity sector
These results shed light upon questions of political economy. Politicians in states like Karnataka or Tamil Nadu are used to thinking that it is difficult to force households to pay for electricity. Politicians in Delhi, Maharashtra and West Bengal have successfully squared this circle: they are able to impose much higher expenditures upon urban households, with the consequential gains for the health of the electricity system and for state public finance.
The most striking finding here is the table organised by consumption quartiles. Why do top quartile households of one group pay Rs.1500 a month for electricity while the same kinds of households of the other group pay Rs.75? This table helps expand the political imagination across the country: What are the politicians of Maharashtra, Delhi and West Bengal getting right, that others are not? There should be a strong demonstration effect here: How are politicians in Maharashtra, Delhi and West Bengal doing the right thing and politically surviving? What is the political settlement in these states that has enabled their superior durable arrangement?
A striking feature of these results lies in the fact that Delhi, which is pooled with Bombay and Calcutta in this work, actually has a large on-budget electricity subsidy program. Even though Delhi has a big subsidy program going, we have a strong result where the group of cities with private distribution includes Delhi. This suggests that private distribution adds value even under a large on-budget subsidy. This brings a new nuance to the debates around the gains from a transparent on-budget subsidy (Jaitly and Shah, 2024). Private distribution, which brings gains in collection and theft reduction, seems to matter over and above the standard public finance gains from a transparent on-budget subsidy.
Conclusion
These results reflect a summary statistic of the working of the electricity system in the two groups of cities, bringing together all aspects that impact household payments for electricity, including overt subsidies, various mechanisms of theft, and the efficiency of bill collection. Future research is required on the sources of improvement through private distribution. Is it better metering and billing technology? A regulatory framework that insulates tariff-setting from short-term politics? Better enforcement against theft?
Section 9.7 of Jaitly et al. 2025 shows the elements of information that go into monitoring the electricity reform at the level of one state of India. The analysis presented here carries this objective one step forward.
This article emphasises the comparison between five cities and not states. Potentially, there can be a divide-and-conquer approach where urban distribution reforms are separated out from the remainder of the state. Once a successful distribution company is established, its footprint can be gradually enhanced, e.g. there is a ready path for the successful private distribution model in Bombay to cover the full footprint of the Mumbai Metropolitan Region (MMR).
Jaitly and Shah, 2021, emphasise that the path to the Indian climate transition runs through the problems of the electricity sector. The political economy of electricity subsidies and theft is a key problem holding back the electricity sector. The empirical political economy results here help illuminate the questions and show pathways to progress.
Such natural experiments, with different parts of the country trying different things, represent the gains that come for the electricity system from the Constitutional scheme where electricity was largely made a state subject. When only one solution is used all through the country, there is reduced experimentation and inferior knowledge. There is much merit in the subsidiarity principle: problems should be placed at the lowest level of the government where they can possibly be placed (Shah and Varma, 2024).
Bibliography
The Price of Power: Costs of Political Corruption in Indian Electricity, Meera Mahadevan, American Economic Review vol. 114, no. 10, October 2024 (pp. 3314–44).
The usefulness of the CMIE household survey data for electricity research in India, Susan Das, Renuka Sane and Ajay Shah, The Leap Blog, 8 May 2024.
Pakistan's quiet solar rush puts pressure on national grid, Economic Times, 16 July 2025.
The lowest hanging fruit on the coconut tree: India’s climate transition through the price system in the power sector, Akshay Jaitly, Ajay Shah, XKDR Forum Working Paper 9, October 2021.
Electricity subsidies are getting better, Akshay Jaitly and Ajay Shah, Business Standard, 26 May 2024.
Electricity reforms in the economic strategy of Tamil Nadu, Akshay Jaitly, Renuka Sane, Ajay Shah, XKDR Forum Working Paper 38, February 2025.
The electricity chokepoint in Tamil Nadu public finance, Charmi Mehta, Radhika Pandey, Renuka Sane, Ajay Shah, XKDR Forum Working Paper 31, February 2024.
India needs decentralisation, Episode 47 of Everything is everything, 17 May 2024.
