Seven States, One River and the 1922 Compact That Promised Too Much Water

Imagine seven families agreeing to split a pie, and writing the split into a contract that binds their grandchildren. Now imagine the baker told them the pie was sixteen slices when it was closer to fourteen. That is the Colorado River Compact in one sentence. It is the most consequential water contract in the American West, and it was signed on a promise that the river never made.

Let me lay out the argument the way it deserves to be laid out. This is a story about a mistake, a very understandable mistake, made by capable people in a hurry. It is also a story about what happens when a federal bargain freezes a guess into law. Conservatives have a view about what to do next, and I will state it plainly at the end: the answer to a scarce resource is not a smarter allocation memo from Washington. It is a price, a property right and a willing seller.

The Terms: What a Compact Is, and What an Acre-Foot Is

Two definitions first, because the argument turns on them.

An interstate compact is an agreement between states. The Constitution addresses it in Article I, Section 10, Clause 3: "No State shall, without the Consent of Congress, ... enter into any Agreement or Compact with another State." So a compact is a treaty-like bargain among sovereign states that becomes binding federal law once Congress consents. That is a feature of federalism, and a good one. The states solved a problem among themselves instead of waiting for a bureau to solve it for them.

An acre-foot is the volume of water that covers one acre to a depth of one foot. It is 325,851 gallons. When the numbers below say "million acre-feet," read "millions of those blocks of water." The shorthand used in the water world is maf, and I will use it.

The Colorado rises in the Rockies and runs well over a thousand miles to the Gulf of California. Its basin touches seven states: Colorado, Wyoming, Utah and New Mexico in the north, and Arizona, Nevada and California in the south. In the early twentieth century the river was wild, flooding in spring and shrinking in late summer, and the people who wanted it were growing fast.

California's Imperial Valley wanted a dependable canal and flood protection. The Upper Basin states feared something more subtle. Western water law runs on prior appropriation: first in time, first in right. Whoever puts water to beneficial use first has the senior claim. California and Arizona were growing faster than Colorado and Wyoming, and the northern states worried that a booming south would lock in the senior rights, leaving the river's own headwaters states with nothing to build on. The slower-growing states had a reasonable fear: that the law of first use would hand the river to whoever dug the ditch first.

The answer was to divide the river by treaty among the states before the rush, so that no one's growth would starve anyone else's future. It was a bargain among states, about property, made in advance. On those terms it was an admirable idea.

The Santa Fe Negotiation and Herbert Hoover's Role

The Colorado River Commission met in Santa Fe, New Mexico, in November 1922, with delegates from each of the seven states and a federal representative as chairman: Herbert Hoover, then Secretary of Commerce under Warren Harding. Hoover was an engineer by training and a negotiator by temperament, and the commission was stuck. Each state's delegation argued from its own interest, and an allocation state by state looked hopeless.

The breakthrough was a change in the unit of division. Instead of carving the river into seven pieces, the commission divided it into two basins. The Upper Basin (Colorado, New Mexico, Utah, Wyoming) and the Lower Basin (Arizona, California, Nevada) would each get an equal apportionment. The dividing point was Lee Ferry, in northern Arizona, just below the mouth of the Paria River. A lawyer from Colorado named Delph Carpenter had championed the compact approach for years, and the negotiation owes much to his persistence. Hoover's contribution was to keep the delegations talking until the structure held. The Compact was signed on November 24, 1922.

The key provisions are short:

• Article III(a) apportions to each basin "the exclusive beneficial consumptive use of 7,500,000 acre-feet of water per annum."
• Article III(b) lets the Lower Basin increase its consumptive use by an additional 1,000,000 acre-feet per year.
• Article III(d) says the Upper Basin states will not cause the flow at Lee Ferry to be depleted below 75,000,000 acre-feet for any period of ten consecutive years.

Do the arithmetic. Two basins at 7.5 maf each makes 15 maf. The Lower Basin's extra 1 maf makes 16. Then, twenty-two years later, a treaty with Mexico added another 1.5 maf, which I will come to. The paper commitments added up to roughly 16.5 maf a year. That is the pie the baker described.

The Mistake: Measuring a River in Wet Years

Here is the problem. The commission relied on stream gauge records from the early twentieth century, and those years happened to be unusually wet. The figure the negotiators had in mind for the river's annual flow was around 16 million acre-feet or a bit more. Later studies of tree rings, which record the river's flow in the width of each year's growth going back centuries, indicate that the long-run average is substantially lower. Estimates of the twentieth-century average from reconstructed natural flow tend to land between roughly 13 and 15 million acre-feet a year, and the drier recent decades have run below that.

I want to be fair to the men who signed. They did not have tree-ring reconstructions of the whole basin. They had the records they had. They were building a framework for a river nobody had yet dammed, and a few decades of gauge data looked like a representative sample. It was not. Honest people with the best available evidence got the central number wrong, and then wrote it into a legal instrument that was designed to last forever.

Here is the lesson, and it is a conservative one. A fixed allocation freezes a forecast. When the forecast proves wrong, the law cannot adjust, because the law is the allocation. Markets adjust. A price is a forecast that updates every day. A compact is a forecast carved in stone.

The Boulder Canyon Project Act, Hoover Dam and Arizona's Standoff

The compact needed to take effect, and it needed a dam. The Lower Basin states needed a way to store and control the river, and California in particular wanted the power and the flood control. Arizona was the holdout. Its leaders feared California would grab the river's water and did not trust the compact to protect Arizona's share. Arizona would not ratify the compact until 1944.

Congress broke the stalemate with the Boulder Canyon Project Act, signed on December 21, 1928. The Act authorized a great dam in Black Canyon, on the Arizona-Nevada line, and the All-American Canal to carry water to the Imperial Valley. It provided that the compact would take effect when six of the seven states ratified it, with California's passage of a limit on its own use as a condition. That put the compact into force without Arizona. The Act also apportioned the lower-basin water among California, Arizona and Nevada, and a famous dispute followed about whether it did so.

Hoover Dam rose in the early 1930s. It was finished ahead of schedule in 1936, and its reservoir, Lake Mead, became the largest in the country. The project was a feat of American engineering and it still delivers power and water to millions. When the right people with the right incentives build something, it gets built; the dam was financed by selling the electricity it generated. Power customers repaid the construction cost. That is a user-pays model, and it is the reason I can praise the dam without reservation. For a companion story of what American builders could do when they wanted to, see We Raised the Empire State Building in About a Year, So Why Not a Power Line.

Arizona v. California, the 1963 Decision

Arizona did not stay quiet. It wanted a project of its own to bring the Colorado across the desert to Phoenix and Tucson, and it feared California's established claims would eat into the river first. The fight went to the Supreme Court as Arizona v. California. The Court appointed a special master, Simon Rifkind, to take the evidence, and the Court issued its decision in 1963.

The holding turned on a question about the Boulder Canyon Project Act. Did Congress, in that law, apportion the lower-basin water among the three Lower Basin states, or did it leave the states to divide the river by prior appropriation and state law? The Court held that Congress had made the apportionment itself, and that the Secretary of the Interior was empowered to allocate the water by contract. The numbers: California would get 4.4 million acre-feet of the lower basin's 7.5 maf, Arizona 2.8 maf, and Nevada 0.3 maf.

The point for our argument: the 1963 decision made the Secretary of the Interior, a federal official, the manager of the lower river. Seven states had made a compact to avoid federal management, and a federal court's reading of the federal dam statute put a federal official in charge. Washington arrived in the end. The ruling also helped clear the political path for the Central Arizona Project, authorized in 1968, which now carries Colorado River water through a canal to central Arizona. As a junior user, Arizona's CAP water is the first to be cut in a shortage. The state that fought hardest for its share holds the shakiest priority. Nobody planned that irony. It is simply what priority dates do.

The 1944 Treaty With Mexico

The river does not stop at the border. The United States and Mexico signed a treaty on February 3, 1944, guaranteeing Mexico 1.5 million acre-feet of Colorado River water each year, with some flexibility for the United States in extraordinary drought. The arithmetic is plain: 7.5 plus 7.5 for the two basins is 15, and Mexico's 1.5 makes 16.5 million acre-feet of promises, before counting the Lower Basin's extra million. The river does not know about our paper. It pours what it pours.

The treaty is also a clean example of the constitutional process working: a foreign commitment, publicly debated and ratified by the Senate, not a surprise inherited from an executive agreement.

The Steelman: Why a Century of Management Is Defensible

The best case for the Law of the River, which is the shorthand for the compact, the Act, the decree, the treaty and the other rules layered on top, goes like this.

First, certainty has value. Farmers in the Imperial Valley and cities in Arizona made billion-dollar investments because the rights were clear. A rule that could change tomorrow would have prevented the capital that built the Southwest.

Second, the Compact avoided a worse fate. Without it, a race to build and claim would have produced litigation without end, with the strongest and fastest state winning.

Third, the system has adjusted, through shortage-sharing agreements and interim guidelines.

I accept all three. Certainty matters, and the compact did prevent a free-for-all. But the defense proves too little. The question in front of us is not whether the compact should have been signed in 1922. It is whether freezing quantities into a federal bargain is the right way to handle a river whose flow cannot be predicted. Certainty about quantities is not certainty about water. A right to 7.5 million acre-feet is a right to a number, and when the river delivers less than the number, the certainty becomes a collision. And note that the "adjustments" are exactly what a market would do continuously, only done in sealed rooms by negotiators, years late, with every state holding a veto.

The Conservative Case for Water Markets

Let me state the premises, in order.

1. Scarcity is a fact, and so is the price signal. When water is scarce, someone must decide who goes without. There are two ways to decide: a committee or a price. A committee allocates by political weight. A price allocates by the value each user places on the water.

1. Agriculture uses most of the river, and much of that water grows low-value crops. The Colorado supplies water for a very large agricultural economy in the Imperial Valley and elsewhere, and a large share of it goes to forage crops such as alfalfa, which feeds livestock. I do not say that to mock a farmer. A farmer's water right is property, and the farmer is not at fault for how the law was written. I say it because it shows the opportunity: a city can pay a farm more for a slice of water than the farm earns from it, and the farm would take the deal and still come out ahead.

1. Voluntary trades beat compelled cuts. Cities, farms and tribes have already leased and traded water in ways that show the model works. A farm that fallows a field for a season in return for payment, a district that pays for a neighbor's efficiency upgrade in exchange for the saved water: these trades are voluntary, they compensate the seller, and they move water from lower to higher use.

1. Federal allocation substitutes a bureaucrat's guess for the owner's choice. If the right is truly property, the owner should be able to sell, lease or bank it, within the physical limits of the system and with protection for third parties. Every restriction on that choice is a subsidy to the status quo.

1. Markets do not require repealing the compact. The compact allocates water among states. Within each state, and across states by agreement, rights can be traded. A price would have signaled scarcity long before the reservoirs fell. Instead, the old rules kept the water cheap for those who held rights and invisible to those who did not.

The objection is that a trade out of a farming county can hurt the hardware store, the school and the tax base. The reply is that those harms are real and should be paid for in the trade, not prevented by banning it. Madison's view that a just government protects property of every kind cuts our way: see Madison Wrote That Government Exists to Protect Property of Every Sort.

What the Founding Says About This

"No State shall, without the Consent of Congress, ... enter into any Agreement or Compact with another State." The Framers put the compact power in the Constitution because they expected states to have shared problems and to solve them among themselves under a light federal check. Federalism is a way of letting fifty governments try different answers; see Federalism Lets Fifty States Experiment, and the Constitution Planned It. A compact is federalism at its best, and the first move in 1922 was right.

Where it went wrong was the afterlife. Every new layer of federal statute, decree and contract shifted authority from the seven states to the Interior Department. The Property Clause is why so much of the West belongs to Washington, as Why Washington Owns Most of Nevada and What the Property Clause Allows explains, and the river's story rhymes with it. Where Washington holds the land, the dam and the contract, it holds the leverage. The same logic of federal management, applied from the other direction, drives the fight over the Snake River dams in Four Dams on the Snake River Make Clean Power, So Why Tear Them Down.

What Reform Would Look Like

Here is the reform agenda in plain terms.

• Clarify and honor the rights, then let them trade. States should make it easy for a rights holder to lease or sell water, including short-term leases that do not endanger a right under use-it-or-lose-it rules. Many Western water codes punish a farmer who uses less water by treating the unused water as forfeited. That is a perverse rule, and conservatives should say so.
• Make the data honest. Plan on the drier long-term average, not the wet decades of the early twentieth century. The Compact's founding error was planning on an optimistic number. Do not repeat it.
• Keep Washington's role small and defined. Run the dams, honor the treaty and the tribal rights, measure the flows. Do not decide which crop deserves water.

Nothing here requires a miracle. It requires trusting people who own water to sell it for what it is worth. The compact was a good idea executed on a bad estimate; when you are not sure of the quantity, write a process instead: a market, a price, a way for owners to adjust.

The case in brief

1. The Colorado River Compact, signed at Santa Fe on November 24, 1922 under Herbert Hoover's chairmanship, divided the river between the Upper and Lower Basins at 7.5 million acre-feet each.
2. The negotiators measured the river in unusually wet years, so the paper rights, plus the 1.5 million acre-feet owed to Mexico under the 1944 treaty, exceeded the river's long-run flow.
3. The Boulder Canyon Project Act of 1928 and Arizona v. California (1963) put a federal official in charge of the lower river, so a compact meant to avoid Washington ended in Washington's hands.
4. Fixed quantities freeze a forecast; prices update it. Voluntary water markets, with payments to affected communities, adjust to scarcity faster than any allocation committee.
5. Plan on the drier long-run average, protect the older tribal and farm rights, and let owners trade.

What you can do: look up your own water provider's source and rate structure, then attend its next public board meeting and ask whether customers can lease or trade water voluntarily.