Nearly half the river’s water goes to growing livestock feed and low-value exports.
by Daniel M. Frey
Forty million people depend on the Colorado River. Some of its largest cities are facing real shortages. Yet nearly half of the water put to human use in the basin goes to growing feed crops such as alfalfa, hay and silage for cattle, with a meaningful share exported overseas.
This is not a hydrology problem. It is a political one.
The misuse of water
Growing alfalfa and hay to feed livestock consumes roughly 46% of all the water put to human use in the basin—more than every city and industry combined, and more than any other single use of the river. In the Imperial Valley, a material portion of alfalfa is exported. Foreign agribusinesses, including Saudi Arabia’s Fondomonte and the UAE’s Al Dahra, have leased Southwestern land to grow water-intensive feed and ship it home to countries that have banned the same crop domestically to protect their own water. Arizona’s governor already canceled state land leases with one of these companies over exactly this issue. We are draining a river that 40 million Americans depend on to grow low-value cattle feed.
This is not even a true market outcome. Between 2017 and 2023, the federal Crop Insurance Program paid $5.6 billion to farmers in the seven Colorado River states for drought-related losses, including more than $2 billion to alfalfa and hay growers alone. Separately, the federal government has committed $1.2 billion over three years under the 2023 Lower Basin conservation agreement to pay some of those same farmers to leave water in the river instead — and under “use it or lose it” water law, that money does not necessarily translate into lasting reduced consumption once the payments end. We are subsidizing the least essential, most water-intensive use of the river while barely funding a lasting transition away from it. These policies ignore urgent human needs and support the continued subsidized irrigation of low-value crops.
Because so much of this crop is exported, converting these acres would not meaningfully affect the U.S. or global food supply. Foreign and domestic buyers would simply source feed from places where water is not scarce. What we would lose is a low-value export that is consuming precious water.
Nobody needs Southwestern agriculture to disappear. Israel farms one of the driest landscapes on Earth and still exports produce because it rebuilt its water use around efficiency instead of volume.
Drip irrigation, which Israel pioneered in the 1960s, delivers water at 70% to 90% efficiency, compared with roughly 40% to 60% for the flood irrigation still common in the Colorado River Basin.
Israel recycles 85% to 90% of its municipal wastewater, supplying more than half its total agricultural demand. Israeli farming also shifted deliberately from low-value, high-water fodder crops to higher-value, lower-water produce such as tomatoes, peppers, dates and avocados. None of this is exotic or new.
Why the status quo persists
So why has nothing changed here? Under prior appropriation law, the oldest agricultural water rights, some dating to the 1800s, are legally senior to every city built afterward, including Phoenix, Las Vegas and Los Angeles. That seniority is the leverage in every negotiation.
The Imperial Irrigation District alone holds senior rights to roughly 3.1 million acre-feet a year — nearly a fifth of the river’s average flow, and on its own equal to Arizona’s and Nevada’s entire combined state allocations. That gives a handful of districts a permanent seat at every basin-wide deal because the law guarantees it. When $4 billion in federal conservation funding was distributed through the Inflation Reduction Act, much of it paid agricultural districts to conserve water they were under no legal obligation to relinquish, thanks to their senior water rights.
This is organized and specific, not a vague farm lobby. The Family Farm Alliance, representing irrigated agriculture across 16 Western states, has testified before Congress repeatedly over the past two decades and led a 2025 push for at least $2 billion in near-term drought funding to sustain current agricultural use, not reduce it.
What should change
None of this is an argument against private property or against farmers and ranchers as people. Water rights in the West are real property, built into land value, farm equity and family inheritance. Any transition worth pursuing must compensate people fairly for what the law actually gave them. Conservation is not agriculture’s job alone. Cities owe real cuts too, through tiered pricing, continued investment in water reuse, limits on ornamental turf and honest accounting of growth in already stressed areas.
But fairness among water users does not mean treating every use as equally necessary. Water for people—for drinking, sanitation and food—comes before subsidized hay exports. That ranking is not radical. It is what “beneficial use” was always supposed to mean before the term became a shield for the status quo.
The technical plan is not the missing piece; versions of it have circulated for years. What is missing is the political will to say, plainly and before the next crisis, that human need outranks irrigated hay, even when the institutions built to avoid that choice are loud, organized and well funded. The reservoirs will not wait for a more comfortable moment. Neither should we.
The 40 million people who depend on the Colorado River should demand that political and business leaders protect their most precious, life-sustaining resource.
Daniel Frey is a Boulder resident, non-practicing CPA and Certified Energy Manager (CEM). He researches and writes on energy and sustainability issues Top photo: A windrower cuts alfalfa for drying prior to baling in the Imperial Valley of California in June 2007. Photo by Timothy Hearsum/AgStock.
- Absurdity of the Colorado River crisis - August 30, 2026
Make sense. Won’t happen. I remember a presentation in 1976 in the Hydrology Department at the University of Arizona by an Israeli graduate student on their drip irrigation. I mentioned that it would pay for itself in 5 years to a farmer in the Eloy Basin (subsidence and earth fissures from mining groundwater). He said his neighbors wouldn’t do it so the water level would decline just as fast. This is like global warming, First papers about greenhouse effect were in the 1800’s. Strong scientific evidence in the 1970’s. Both reservoirs were full in 2000. Climate change and drought have been coming for us slowly. Upper Basin gets no mandatory cuts. Lower Basin gets mandatory cuts. No article says whether cuts come from the 7.5 maf allocation or current usage of 6 maf. Upper Basin claims to have been using only 4 maf for a long time. Total flow is supposed to be 10 maf. Until the last couple of years, Lees ferry has had flow of more than 7.5 maf. An additional 1 maf at the Diamond Creek gage into Mead. The numbers do not add up with 1.5 maf evaporative and infiltration loss and depletion in the two reservoirs. We need a reach by reach plot of discharge in maf/yr of the whole system to see where the water is coming from and going to. I think only Flaming Gorge of the federally funded Upper Basin reservoirs has been depleted at all to move water to Lake Powell. I think it is very possible that some users are taking more than they are allowed.
Provocative essay, but he doesn’t dig into the nuances of consumptive versus non consumptive use nor does he acknowledge the greenway value of irrigated hay meadows. The argument seems to be, “Let’s get rid of stupid alfalfa because only meat eaters need it.” (He didn’t say that—but my crystal ball picked it up.)
With the catastrophic climate conditions of especially the last two years, the only available forage for wildlife in northwestern Colorado affected by drought and fire, is the irrigated river bottom which typically returns, according to the state water engineer roughly 80% of the water used to the river. Google earth aerial photos illustrate this well.
The bottom line is this: water rights are irrelevant when there is no water. The 30 year precipitous decline in available water is the issue. When will we step up—not to stop climate change—but to begin the effort to reverse it?
The exportation of hay is another matter and I think we can agree on that one whole-heartedly.
I agree with the sentiment — we have a water management crisis, much more than a water crisis. But let’s also talk about the stupidity of unbridled urban growth in some of hottest and driest places in the United States.
Eminent Domain. To help completely retire the senior rights folks like the Imperial Valley have.
I don’t mean the federal government should condemn every alfalfa farm in the Imperial Valley, take possession of the dirt, and become the world’s largest and least competent farmer.
What I’m wondering about is something more targeted: why couldn’t Congress create a permanent Colorado River water-retirement program, fund it adequately, and ultimately use condemnation as the backstop when voluntary acquisition isn’t sufficient?
In other words, don’t move the water from alfalfa to subdivisions, golf courses or data centers. Permanently remove a defined quantity of consumptive use from the system and leave the water in the river/reservoir system.
The federal government has used eminent domain on an enormous scale when it decided that highways, dams, reservoirs, military installations, parks and other infrastructure were national priorities. The Interstate Highway System is the obvious example: the government didn’t say the existing property-rights regime made a national highway network impossible. Congress authorized the project, property was purchased wherever possible, and condemnation existed when voluntary acquisition failed.
Why couldn’t we apply some version of that thinking to the Colorado River?
For example, Congress could decide that the system needs to permanently eliminate, say, 1–2 million acre-feet of annual consumptive use. Reclamation could identify where permanent retirement produces the most actual water per federal dollar, with safeguards for communities and the environment.
Then:
1. Offer generous voluntary buyouts first — probably well above conventional agricultural land values because we’re really buying the consumptive water use, not merely dirt.
2. Purchase or condemn whatever legally cognizable property interests are necessary — water rights, contractual entitlements, easements, land where necessary, etc. — with full Fifth Amendment compensation.
3. Permanently retire the associated consumptive use. The conserved water does not become available for Phoenix, Las Vegas, Los Angeles, golf courses or data centers. It stays in the system.
4. Target the program by water consumed rather than by crop. I agree with you that simply banning alfalfa makes little sense if the farmer replaces it with another equally thirsty crop.
5. Protect small farmers and communities. Make the voluntary offer especially attractive to small operators and, if compulsory acquisition ever becomes necessary, concentrate it where the largest volumes can be retired with the least economic and community disruption. Congress could also fund transition assistance, local tax-base replacement, Salton Sea mitigation, habitat restoration, etc.
What makes me think this isn’t completely crazy is that the federal government is already doing a temporary, voluntary version of it.
Reclamation has been paying hundreds of dollars per acre-foot for “system conservation” specifically so that wet water remains in Lake Mead. The Imperial Irrigation District agreement alone contemplated up to 700,000 acre-feet of conservation from 2024–26 for roughly $589 million.
So we’ve already accepted the principle that taxpayers can compensate existing entitlement holders specifically for NOT consuming Colorado River water.
My question is: Why keep renting the same water?
If the structural deficit is permanent, wouldn’t it ultimately make more sense for the federal government to buy and permanently retire enough consumptive use to bring demand into rough balance with the river we actually have?
And if 95% of the necessary rights can be acquired voluntarily, great. Eminent domain isn’t Plan A. It is what prevents the final 5% of holdouts from making a congressionally authorized basin-wide solution impossible.
I also think this addresses your objection to Yglesias. I agree that transferring an Imperial Valley water right to a subdivision doesn’t necessarily solve the Colorado River problem. That’s just changing who consumes the water. What I’m proposing is extinguishing or permanently retiring the consumptive use itself.
Maybe existing Western water law makes that harder than I’m imagining. That’s actually what I’m curious about. If Congress explicitly authorized Reclamation to acquire and permanently retire enough Lower Basin consumptive use to produce a specified amount of system water, what is the legal obstacle that prevents it?
Because at some point it seems strange to me that we’re willing to contemplate the collapse of Powell and Mead before we’re willing to contemplate paying fair market value to retire property rights that were created around a river that no longer produces enough water to satisfy them.