The Groundwater Bill

A farmer in central Punjab who wants to irrigate a hectare of paddy turns on a pump and pays nothing extra. Not "very little" — nothing. His electricity bill was settled by the size of his pump, in horsepower, months ago. The water that comes out of the ground this afternoon is free at the margin, because there is no meter between the transformer and the motor, and the tariff is not built to need one.

That is the whole mechanism, and it is an accounting decision rather than a moral failing. Price the last cubic metre at zero, guarantee a buyer for the crop that responds best to water, and you have instructed several million people to pump as much as their land and their electricity supply allow. They did. The aquifer is the account that received the instruction.

The pump has no meter, and that is the design

Canal irrigation is measured. Water arrives on a roster, through an outlet sized in cubic feet per second, and a warabandi schedule tells each farmer which hours are his. Tubewell irrigation is not measured at all. The connection is billed on connected load — a fixed charge per horsepower per month, often waived outright in the paddy states — and the marginal cost of running the pump is the diesel or the units that the subsidy already covers.

This is why the usual advice, "charge farmers for water", does not translate from other countries. There is nothing to charge against. The irrigation department has no device at the wellhead, the power utility does not read a farm meter because farms mostly do not have one, and the pump itself is a private asset on private land. In the canal commands, the state at least owns the gate.

The consequence is a price signal that runs exactly backwards. A farmer with a fixed electricity bill and a guaranteed procurement price faces a marginal cost of zero for water and a positive marginal revenue for every extra quintal. The rational response is to keep the field flooded, to plant the water-hungry crop, and to deepen the borewell when the column drops. Each of those decisions is individually sensible. The sum is not.

Why the water leaves faster than it arrives

Punjab and Haryana sit on alluvium, which is why the Green Revolution was possible there: deep soils, high storage, recharge from canal leakage and from the monsoon. Alluvium is also why depletion took decades rather than years. A hard-rock aquifer in the Deccan gives up its water quickly and answers a drought with a dry bore; a thick alluvial aquifer can be mined for a generation before anyone's pump fails, which delays the feedback that would otherwise stop the behaviour.

The monsoon is the other half of the account. Recharge is not a salary paid every month; it is a lump that arrives, or does not, in three months. A year of normal rain refills part of what a year of double-cropping takes out. A weak monsoon year does not merely fail to refill — it triggers extra pumping, because the canal roster is short and the crop still needs water. The deficit ratchets.

What makes this specific case hard is the combination: free power at the margin, an assured buyer for paddy, a cropping calendar that runs two crops a year on the same field, and a procurement system that will take every kilogram that arrives. Remove any one and the arithmetic softens. All four have been in place together for half a century.

A well is private property drawing on a common pool

The pump belongs to the farmer. The water under it does not, legally, belong to anyone in particular — it belongs to the landowner above it, in practice, with no limit on how much can be taken and no register of who took what. That is the quiet legal fact behind the hydrology. Deepening a borewell is a private investment that yields a private return and draws down a shared stock. There is no moment at which any single farmer faces the cost of the water he removes from everyone else's column.

Wells also become businesses. Where a farmer has surplus pumping capacity he sells water by the hour to neighbours with smaller holdings, which turns the subsidy into an income stream and gives the biggest pumps a reason to keep drawing. The informal water market is not a distortion on top of the subsidy; it is one of the ways the subsidy is converted into cash.

The state pays twice

The subsidy is usually counted once, as a line in the power utility's books. It is really two payments. The first is the cash: the state reimburses the utility for the electricity it gave away, several thousand crore rupees a year, which is money not spent on canals, drainage or storage. The second is the water, which is not in any budget, because a depleted aquifer has no owner to send an invoice.

There is a third cost, quieter. When the table falls, the shallow wells fail first. The farmers who cannot afford to drill deeper are the ones who leave tubewell irrigation, and the ones who can afford it buy more land or sell water to their neighbours. A subsidy that is uniform per horsepower is, in its effect, regressive: it is worth most to the farmer with the biggest pump.

The one measure that worked, and how much it worked

In 2009 Punjab passed an act that did something unglamorous and effective: it pushed the transplanting of paddy later, to coincide with the arrival of the monsoon, and restricted nursery raising before a notified date. The point was not to save water per kilogram but to stop pumping during the six weeks when evaporation is at its peak and the rains have not yet come.

Satellite and well-hydrology studies since then report that the rate of decline slowed in the central districts. Slowed, not reversed. That is the honest result, and it is instructive: a single well-designed administrative lever moved a trend that thirty years of tariff debate had not touched. It worked because it did not require anyone to pay for water — it changed the calendar instead of the price.

The same logic explains the alternatives on the table. Metering farm power and paying the subsidy as a direct transfer per hectare would restore a price signal, but it requires meters, feeders and a political decision to convert an untargeted subsidy into a targeted one. Shifting area out of paddy into maize or pulses needs a buyer for those crops as reliable as the paddy buyer. Managed aquifer recharge — the recharge shafts and check dams that several districts have built — works where the geology allows it and where the water is not already being pumped out faster than the structure can put it back.

What the bill actually says

The bill is not a bill for water. It is a bill for a way of buying political peace in the countryside and cheap wheat in the cities at the same time, paid in a currency — the water table — that does not appear in the fiscal deficit. That is why it has been so durable: every year the payment is invisible, and every year the balance is a little smaller.

The last honest thing to say is that the depletion is not uniform and not irreversible everywhere. Some blocks have stabilised; recharge structures have lifted local tables; a shift in cropping has happened in pockets without any policy at all, driven by labour costs. The direction of the state-level number is down. A block-level map shows a patchwork, and that patchwork is where any real policy has to start — because the unit that can be governed is the watershed, not the state.