In 1975-76, India's public canal systems recovered 96% of what it cost to run them — not build them, just keep them operating — from the fees charged to farmers. By 1997-98 that had collapsed to 5.8%. It recovered to roughly 20% by 2013-14, then averaged only about 7% across 2000-01 to 2016-17 once the full run is counted. Nobody has published a national figure past 2017-18. This is the public, canal-and-dam half of India's irrigation story; the private half — the 96.6% of India's 23 million irrigation schemes that are borewells and tubewells farmers paid for themselves — this blog has already covered. This piece is the other half: what the public money actually gets back.
India's Irrigation System Recovered 96% of Its Running Costs in 1975. It Recovers About 7% Now.
A generous measure, and it's still failing
"Financial recovery" for public irrigation in India has a specific, narrow meaning, and it's worth stating before the number, because the number looks bad even under the most forgiving version of the question. The Central Water Commission (CWC) defines it as gross receipts — water-fee collections and other charges — divided by working expenses, meaning operation and maintenance costs alone. It deliberately excludes the original fixed capital investment: the canal, the dam, the distributary network the state already spent to build. This is the recovery rate on running the system, not on having built it. Even by that generous definition, gross receipts covered only about 7.37% of working expenses on average between 2000-01 and 2016-17 — split between 1.59% recovered from actual water-sale fees and 5.79% from other charges. The system, on its own terms, doesn't cover the electricity and staff bill to keep the canals flowing, let alone anything toward what it cost to dig them.
The collapse, and the partial recovery, in one series
The CWC's own historical figures trace a specific arc, not a flat failure. In 1975-76, gross receipts covered roughly 96% of working expenses — irrigation fees were doing close to the job they were designed for. By 1997-98 that had fallen to a mere 5.8%. It partially recovered to about 20% by 2013-14, before the fuller multi-year average (2000-01 to 2016-17) settles at the much lower 7.37% figure above — a reminder that a single good year sitting inside a longer bad stretch can flatter a headline if quoted alone. One counterintuitive detail worth keeping: research on this data has found the recovery rate tends to run higher in India's less economically developed states than in its more developed ones — the opposite of what a simple "richer states manage their utilities better" prior would predict.
| Period | Recovery rate (gross receipts ÷ working expenses, %) |
|---|---|
| 1975-76 | ~96 |
| 1997-98 | 5.8 |
| 2013-14 | ~20 |
| 2000-01–2016-17 average | 7.37 (1.59 water fees + 5.79 other) |
Source: Central Water Commission data as analysed in the Springer chapter "Financial Performance of India's Irrigation Sector: Past, Present and Future," part of the volume this piece's sources note names in full below. The CWC's own methodology counts only operation-and-maintenance working expenses against gross receipts; it does not net against, or even mention, the fixed capital cost of construction.
What the spending side adds — and what this piece can't add to it
This blog's own reporting on India's Minor Irrigation Census already introduced a 2024 peer-reviewed paper relevant here — Bathla, Kannan & Das, "Public investment in irrigation across the Indian states: Financial recovery and governance" (Agricultural Economics Research Review 37(1): 93–112) — in the context of that piece's private-irrigation argument. This piece is the public-investment counterpart the earlier one flagged but didn't develop: the paper examines public capital formation in major, medium and minor irrigation across 20 states from the triennium ending 1983-84 to the triennium ending 2019-20, alongside a governance index built from state-level indicators for 2001-02 to 2015-16. Its own published abstract states that a sizeable increase in public expenditure on irrigation during the 2000s "has barely corresponded" with net irrigated area, potential utilised, or the rate of financial recovery, and that recovery itself is positively associated with a state's agricultural income, capital stock, and area under rice and wheat — meaning the states that recover the most are the ones already concentrated on the two crops this blog's own urea and cropping-pattern coverage keeps returning to.
The gap persists into the current year
The spending-versus-outcome mismatch the 2024 paper describes is still visible in the latest available national figures. As of 2024-25, India has created irrigation potential of roughly 112 million hectares against an ultimate irrigation potential estimated at about 139.9 million hectares — but actually utilises only about 90 million hectares of what's been created, a gap of roughly 22 million hectares that is created, funded, built, and not being used to irrigate anything. This is a persistent, named structural problem in India's own irrigation planning literature, not a one-year anomaly; the same gap between potential created and potential utilised has been flagged as a "major challenge" across successive irrigation reviews.
The current flagship spending vehicle shows the same shape of gap on its own numbers. Pradhan Mantri Krishi Sinchayee Yojana (PMKSY) was continued for 2021-26 with an overall outlay of ₹93,068.56 crore; as of August 2025, ₹64,407 crore had been released, reported as benefiting more than 27 million farmers through irrigation, micro-irrigation, watershed and groundwater components since the scheme's 2016-17 launch, with a further ₹6,587 crore budgeted for FY2026-27. Within that, the Accelerated Irrigation Benefits Programme (AIBP) component — the part that funds specific named dam and canal projects — had completed 71 of 123 projects (57.72%) as of 8 May 2026, and created 2,940.97 thousand hectares of irrigation potential against a target of 4,531.86 thousand hectares (64.90%). Potential-created is running ahead of project-completion, which is normal (a project can create partial potential before its last component is finished), but both figures sit meaningfully under 100% on a scheme now a decade old. The programme keeps adding to its own project list even as the existing backlog works through: the Cabinet approved adding Uttarakhand's Jamrani Dam and Bihar's Kosi-Mechi link to AIBP's coverage in 2026, each a multi-year addition to a completion rate that hasn't yet cleared 60%.
What doesn't follow from any of this
None of the above shows that Indian irrigation delivers no value for the money spent on it — a canal system can generate large, diffuse agricultural and food-security benefits that never show up as a fee collected from a farmer, and a recovery rate measured against water charges alone was never designed to capture that. Nor does the historical 96%-to-7% arc mean recovery has fallen in a straight line; the 2013-14 partial rebound to roughly 20% shows the trend isn't monotonic, and a multi-year average can sit well below even a recent good year. This piece also does not establish why less-developed states recover a higher share than more-developed ones, only that research on the data has found this pattern; the mechanism is not stated in anything this piece could access. And the "22 million hectare" utilisation gap is a national aggregate; this piece does not have a state-by-state breakdown of where that gap actually sits.
What would fix this
- Publish the recovery-rate series on a fixed annual schedule, not a three-to-five-year lag. A number that's structurally always years stale can't inform a current budget decision, only describe a past one.
- Report recovery net of, not excluding, fixed capital cost, alongside the working-expenses-only figure. The current methodology is the most forgiving version of the question by design; publishing the stricter number next to it would show how much of the gap the generous framing is currently hiding.
- Publish the state-by-state breakdown behind the "less-developed states recover more" finding. If that pattern holds up to scrutiny, it complicates a lot of assumptions about which states manage irrigation well; if it doesn't, the aggregate claim shouldn't keep circulating unexamined.
These are proposals from this blog, not recommendations any agency has adopted.
Sources and caveats
Irrigation cost-recovery series (96% in 1975-76, 5.8% in 1997-98, ~20% in 2013-14, the 7.37%/1.59%/5.79% 2000-01–2016-17 averages, the CWC's working-expenses-only methodology, the 3-to-5-year reporting lag, and the less-developed-states finding) per the Springer chapter "Financial Performance of India's Irrigation Sector: Past, Present and Future," in The Future of Irrigation in India (2022, DOI-linked at link.springer.com/chapter/10.1007/978-3-030-89613-3_7), itself analysing Central Water Commission data; this piece worked from search-indexed summaries of that chapter and the related peer-reviewed article "Financial performance of India's irrigation sector: a historical analysis" (International Journal of Water Resources Development, vol. 34, no. 1), since link.springer.com, tandfonline.com and cwc.gov.in were all unreachable to direct fetch from this piece's research environment today. Bathla, Kannan & Das, "Public investment in irrigation across the Indian states: Financial recovery and governance" (Agricultural Economics Research Review 37(1):93–112, 2024) is cited from its own public abstract only, as it was cited in this blog's own earlier reporting linked above; the full text (epubs.icar.org.in and a ResearchGate mirror) returned a network-egress block on every attempt made for this piece. Current irrigation potential-created/utilised figures (112M ha created, ~139.9M ha ultimate potential, 90M ha utilised, ~22M ha gap) per Down To Earth's reporting, itself citing government irrigation-planning data; this figure was not independently re-verified against a primary Ministry of Jal Shakti table for this piece, and should be treated as a widely-reported secondary figure rather than a directly-read primary one. PMKSY's 2021-26 outlay (₹93,068.56 crore), the ₹64,407 crore released as of August 2025, the 27-million-farmer figure, the FY2026-27 ₹6,587 crore allocation, and the AIBP completion figures (71 of 123 projects, 57.72%; 2,940.97 of 4,531.86 thousand hectares, 64.90%, as of 8 May 2026) are consistently repeated across multiple secondary sources (NextIAS, IBEF, Lukmaan IAS) that this piece could not trace back to a single named PIB press release ID or Lok Sabha answer number despite trying; treat these as reliable-by-repetition rather than directly verified against a primary document. The Jamrani Dam and Kosi-Mechi Cabinet approvals are each named, dated PIB press releases, but pib.gov.in itself returned a network-egress block on direct fetch today, so this piece cites their content as reported by secondary coverage rather than the release text itself. Today's research environment blocked an unusually wide set of domains on direct fetch (including several that have been reachable in this blog's past research), which is disclosed here rather than papered over; readers who need to independently verify the CWC recovery-rate series or the PMKSY/AIBP figures should go to the Central Water Commission's and Ministry of Jal Shakti's own published reports directly. Compiled via web research on 25 September 2026; nothing here is investment, agricultural or policy advice.
About this article: Researched, written and edited by Umashankar Triplicane Dwarakanathan, with AI research assistance; every figure is meant to trace to the primary source cited. See the Editorial Policy for how sourcing, AI use and corrections work.
