The Water You Cannot See Moving

Three rivers, three disputes, and a question that should have been asked at the start: how much water actually crosses borders as water?

A surprisingly small share. Most of it crosses as food.

The reason is simple once stated. Growing a crop consumes water where the crop grows. Raising an animal consumes the water that grew its feed. When the product is sold abroad, the buyer has obtained the use of water that fell somewhere else, without any of it moving in a pipe. John Anthony Allan, working on the Middle East, named this virtual water in the early 1990s and was awarded the Stockholm Water Prize in 2008 for the idea. The prize's own illustrations are useful for scale: roughly a hundred and forty litres behind a cup of coffee, something like two thousand four hundred litres behind a hamburger.

The policy implication was immediate and counterintuitive at the time. A water-scarce country's cheapest water project may be a food import programme. Buying wheat costs less than irrigating it, uses no aquifer, and arrives by ship. Several Gulf states concluded exactly this and wound down ambitious domestic cereal farming that had been draining their groundwater.

That is where the Arizona story comes in, and it is worth telling precisely because nothing in it is secret or illegal.

Fondomonte, a subsidiary of the Saudi dairy group Almarai, farms alfalfa in the American southwest and ships it to feed dairy herds in Saudi Arabia. Alfalfa is a thirsty crop. Some of this farming took place on Arizona state land leases in Butler Valley, in a rural basin where groundwater pumping was effectively unmetered and unlimited by the terms of the lease. In October 2023 the state terminated one of those leases, citing default, and announced it would not renew three others due to expire the following February.

Set the two halves of this book side by side.

The Nile dispute has a named structure, a date of inauguration, a complaint to the Security Council and a decade of ministerial negotiation. The Indus dispute has a treaty, a commission, neutral experts and an award from a court in The Hague. The Mekong dispute has a commission, a rival forum, satellite studies and a data-sharing agreement. All three are visible, documented and argued about by foreign ministries.

The transfer of water embedded in fodder from a desert aquifer in Arizona to a dairy herd in Saudi Arabia was governed by a state land lease, a commodity contract and a shipping schedule. No ministry had jurisdiction over it as a water question, because on paper it was not one.

This is not an argument that the trade is wrong. Virtual water flows are, in aggregate, why food-importing dry countries can feed themselves at all, and they move water use towards places where water is abundant more often than they move it away. The argument is narrower and harder to answer: the governance follows the visibility rather than the volume. Rivers get treaties because rivers are legible. Aquifers get lease renewals.

If you want to know how much water a country controls, the trade statistics will tell you more than the river map. And the single most useful habit this book can leave you with is the question of who is bearing a cost that no document in the transaction records.