Articles · Note on access, scarcity, privatisation and the persistence of a failed reform model
Draft National Electricity Policy 2026: a critical evaluation
EEFI Centre
The draft says access and scarcity are solved, then tightens the same model: cost-reflective tariffs, an exit from universal service, capacity markets, and public money under private assets.
The Draft National Electricity Policy, 2026 presents itself as the map to reliable 24×7 power for 2047. Its first political act is historical. Access and scarcity, it says, are solved. What remains is DISCOM finance, industrial tariffs, and a green transition that private capital and markets will deliver. That is why the draft pushes cost-reflective tariffs, exemption from Universal Service Obligation for large consumers, surcharge-free open access for manufacturing, capacity markets, competitive bidding inside the states, listing and public-private partnership in distribution, private nuclear power, and public money to de-risk private finance.
Those founding claims have to be tested first. A village has been “electrified”, since 2004, if there is a transformer and lines in the inhabited area, electricity in the school, the panchayat and the health centre, and at least 10 per cent of households connected. All inhabited census villages were declared electrified by 28 April 2018 on that definition. That is some infrastructure on paper. It is not a connection in every home, and it is not usable supply around the clock. Saubhagya counted anyone who did not apply as unwilling. After the scheme closed, a household still needs identity, occupancy proof, in most states an installation report from a licensed contractor, and internal wiring beyond the single light point the scheme itself provided. The free service line is short. Distance, poverty and an unwired house are written down as lack of willingness.
The 100 per cent figure was made by shrinking the target. The guidelines spoke of about 300 lakh unelectrified households; an earlier figure in the same guidelines was 517 lakh; the dashboard later used 248.48 lakh. CAG Report 17 of 2025 found that “100 per cent” coverage by March 2019 was declared after that revision. Actual Saubhagya connections by March 2019 were about 151.60 lakh — under 58 per cent of the original 300-lakh target. The Revamped Distribution Sector Scheme has since sanctioned grid works for another 13.65 lakh households among particularly vulnerable tribal groups, tribal, Scheduled Caste and border families. NFHS-5 put household electricity at 96.8 per cent, not 100. Access improved. The claim that access is closed is a definitional victory.
Small annual energy-deficit percentages are not abundance. The CEA’s Load Generation Balance Report for 2026–27 still projects a Southern peak deficit of about 9.6 per cent and an Eastern peak deficit of about 16.2 per cent, beside a thin all-India surplus on paper. In April and May 2026 the grid fell short on 36 of 61 days at the non-solar-hour peak, and on only 6 of 61 days in solar hours. After a weak monsoon, September 2026 night shortfalls reached about 7.7 GW. From 6 pm to 11 pm the day-ahead price sat on the ₹10 ceiling. On stressed evenings, purchase bids ran far ahead of sell bids. The High-Price Day Ahead Market exists so that gas and imported-coal plants can sell above the cap. Coal at the plant is a coordination failure: mid-September 2026 thermal stocks were about 7 to 9 days, a three-year low, while mine-head inventories could look comfortable. Coal, Railways and Power each chase their own commercial target.
As on 31 March 2026 the official split was Central 23.3 per cent, State 24.4 per cent and Private 52.3 per cent. Private plants run when the power purchase agreement or the exchange pays. They have no social duty to dispatch. Capacity payments, the high-price market, and a ceiling with an escape hatch turn tightness into revenue. Capacity rose. Evening scarcity, coal failure at the plant gate, and exchange prices stuck at or above the cap remain.
The draft treats distribution as the leftover failure of 2003, and points to losses, tariffs and monopoly. Aggregate technical and commercial losses exist. They are not the main fiscal engine. Power purchase is 70 to 80 per cent of DISCOM cost. Generators recover two-part tariffs, change-in-law claims and late-payment surcharge. Dues to generating companies fell sharply after the late-payment rules. That is a shift of liquidity toward generation, not proof that purchase prices fell. Accumulated losses are about ₹6.5 to 6.9 lakh crore. Outstanding debt on 31 March 2025 was ₹7.26 lakh crore. A small all-India profit after tax in 2024–25 is private profit of about ₹12,138 crore pulled across state DISCOM losses of about ₹9,437 crore. Regulatory assets — this year’s unpaid cost of supply, parked for a future tariff shock — run from about ₹1.74 lakh crore to near ₹3 lakh crore, concentrated in Tamil Nadu, Uttar Pradesh and Rajasthan. Bailouts have followed one another since 2001. Debt and regulatory assets rebuild after each clean-up. Generation was unbundled on cost-plus terms. Retail tariffs stayed politically capped. Paying industry was invited to leave through open access.
Read as a whole, the draft does four things, and they are one design. It rewrites history, so that what follows looks technical. It completes the unfinished business of 2003 in the segments still under public and state control: intra-state transmission, the distribution wire, bulk storage, and nuclear power. Consumers of 1 MW and above may leave the Universal Service Obligation. Manufacturing, Railways and Metro are to be freed of cross-subsidy and additional surcharge. It socialises risk and privatises return: viability-gap funding, first-loss covers, green bonds, recovery of old regulatory assets through future tariffs, and capacity payments for availability. The draft’s own financing figures — ₹50 lakh crore by 2032 and ₹200 lakh crore by 2047 — are not a plan of public investment. The state supplies predictable revenue and takes the first loss. Private capital takes the asset.
A public infrastructure alternative is not obscure, and the draft does not take it. Public and publicly owned cooperative investment in renewables, storage, intra-state grids and nuclear power. Universal service for every consumer, with no exit above 1 MW that does not compensate the common pool. Distribution as a publicly owned, socially accountable service. Planning that sets the generation mix, and social dispatch rather than scarcity pricing. Public banks for public assets. Agri-feeder solarisation as a programme for farmers, not only as a device to cut subsidy. Coal, rail and power run as one chain with priority for the power sector. Access defined as a wired, affordable, adequate household connection — not as 10 per cent of a village plus willingness. The full note, section by section, is attached.