Three Rational Decisions That Killed Owens Lake

What the System Looked Like

The Owens River gathered Sierra Nevada snowmelt and carried it south through a valley roughly a hundred miles long. The Sierra walled it to the west. The Inyo and White Mountains walled it to the east. The valley floor was high desert that irrigation had softened into something workable, and the towns strung along the river and the railroad in the order you'd expect: Bishop at the north end, then Big Pine, Independence, Lone Pine. At the basin's southern terminus sat Owens Lake, saline and shallow and alkaline, collecting runoff from the surrounding mountains and losing it back to the atmosphere. No surface outlet. The dissolved minerals had been accumulating for millennia. The cycle was old and stable and nobody thought of it as a cycle. The lake reflected the Sierra when the wind was still.

A 1910 federal agricultural census inventoried what the water supported: 43,000 sheep, 20,000 cattle, 5,000 horses. Twenty thousand apple trees. Forty thousand grapevines. Fifty-eight thousand bushels of corn, fifty-one thousand of wheat, fifty-three thousand of potatoes. A hundred tons of honey from 5,800 bee colonies working the irrigated orchards. The valley was not paradise. A visiting horticultural specialist found farmers flooding unleveled land, contributing to over-irrigation and alkalinity. Railroad access was poor. Freight charges were high. But the water came every year, and the river ran, and the lake sat at the bottom of the basin doing what terminal lakes do: collecting, evaporating, persisting. It was the endpoint of a system that functioned, and nobody was tasked with maintaining it, because it maintained itself.

By 1933, Los Angeles owned 85 percent of the valley's residential and commercial property and 95 percent of its farm and ranch land. Irrigated acreage dropped from roughly 75,000 acres in 1920 to 23,625 by 1940. Most of the purchased land reverted to semi-arid desert. The towns shrank to what the National Park Service called "semi-ghost towns." The lake was gone.

Between those two snapshots, three decisions made between 1903 and 1907 operated as a ratchet. Each narrowed the valley's range of possible futures. Each was rational within its own frame. Together they locked in an outcome that was irreversible before most of the people it would destroy understood what was happening.

The Buyer

The idea predated the buying campaign. Frederick H. Newell flagged Owens Valley as a potential reclamation project in spring 1903, and the Reclamation Service sent a surveyor to study feasibility. The proposed federal irrigation project included a 260,000 acre-foot reservoir and canals to irrigate thousands of acres, potentially serving as many as 185,000 acres. This was the valley's plausible alternative future: federal investment in local agriculture, exactly the kind of project the 1902 Reclamation Act had been designed to build. The water stays. The valley grows. The lake persists.

Fred Eaton, former mayor of Los Angeles, was reading the same water from the other end of the pipe. He had a longstanding vision for bringing Owens Valley water south, and his relationship with William Mulholland, the city's water superintendent, gave the vision institutional reach. Mulholland had concluded by 1904 that local Los Angeles water sources were inadequate for future growth. He'd surveyed rivers and groundwater basins south of Tehachapi and the numbers told him what numbers tell you. The Owens Valley aqueduct became his answer: gravity-flow alignment, aqueducts and reservoirs, no pumping required. Two hundred and thirty-three miles of downhill. Eaton was the man who knew the valley. Mulholland was the man who could build the pipe.

Eaton arrived in Owens Valley in March 1905 and started buying. He paid $450,000 for a two-month option on ranch lands and 4,000 head of cattle. He acquired rights to more than fifty miles of riparian land, including the Long Valley reservoir site. The Los Angeles Times later reported that he had purchased the Rickey Cattle Company holdings, "comprising about 50,000 acres of water-bearing land," and that it had been thought Eaton was entering the stock business.

What Eaton told individual sellers about his purpose is where the record fractures. LADWP's own institutional history acknowledges he visited the Independence Land Office during the same trip to do research for J.B. Lippincott, the Reclamation Service's California-region engineer, and that "whether deliberate or not," he left the impression he was doing Reclamation Service work. A 1907 San Francisco Call report said Eaton came to the Independence land office around March 1, 1905 with a Lippincott letter directing him to obtain information and examine power projects, while simultaneously buying riparian lands and water rights later turned over to Los Angeles. Valley land register Stafford Wallace Austin and receiver Richard Fysh accused Eaton of misleading sellers. A 2024 legal scholarship review argues the evidentiary basis for deliberate deception is "remarkably thin" and that the story "morphed continuously" through summer and fall 1905. The same review notes that what one account calls secrecy can also be read as ordinary acquisition practice: announcing a water-rights buying campaign would have caused sellers to raise prices.

The sellers had reason to believe they were participating in something that would benefit the valley. A federal reclamation project was under active study. A man carrying a Reclamation Service letter was asking questions at the land office. Some sellers thought they were selling to a cattleman. Others may have believed they were facilitating the federal project. The difference between selling land for a hoped-for irrigation project and selling water rights to a distant city was the difference between a future with water and a future without it. By the time the distinction became clear, the options were signed.

Whether Eaton lied matters morally. Structurally it changes nothing. By the time the buying campaign became public, he had secured options on what LADWP describes as "basically all parcels of any importance not controlled by the Reclamation Service." The key reservoir site and fifty miles of river frontage were committed to Los Angeles before the valley understood the commitment.

And then there was Lippincott. The Reclamation Service's own California-region engineer had signed a March 1905 contract with the Los Angeles Water Commission to prepare a comprehensive survey of possible city water sources while still holding his federal position. The degree of collusion between Lippincott and Los Angeles is disputed. Abraham Hoffman's Vision or Villainy argues that Roosevelt-era utilitarian views across the Geological Survey, Reclamation Service, and Bureau of Forestry helped determine the outcome. William Kahrl's Water and Power reads the evidence as more damning. What is not disputed: the man responsible for evaluating whether the valley deserved a federal irrigation project was simultaneously under contract to the city that wanted the valley's water. He was reading the same system from two positions and billing both.

The Reclamation Service's Board of Engineers met on July 27, 1905, and found the Owens Valley project viable from an engineering standpoint. But the board also noted that Los Angeles's ownership of Long Valley and fifty miles of river frontage was now a major impediment. The project was abandoned as Los Angeles continued acquiring land and water rights. The valley's alternative future was foreclosed by prior acquisition, not by any failure on its merits. The land was already committed before the evaluation was complete, and the evaluation found the commitment was the problem.

The Vote

On July 29, 1905, the Los Angeles Times broke the aqueduct plan under the headline "Titanic Project to Give City a River." Thirty thousand inches of water from the Owens River through a 240-mile conduit. Enough for a city of two million. Mulholland was quoted the same day: "The last spike is driven…the options are secured."

Mulholland had spent years working his way up through the Los Angeles water system, from the labor crews to the superintendent's office. He understood the city's water infrastructure the way a field tech understands it: from the physical system itself, the capacity of every pipe and reservoir, the seasonal curves, where the pressure dropped and why. When he said local sources were inadequate, it wasn't speculation. He had measured them. The Owens Valley aqueduct was his solution, and by the standards of early twentieth-century engineering it was elegant. Two hundred and thirty-three miles of gravity-fed conveyance, no pumps. Mulholland was not a grifter. He was an engineer who had identified a real constraint, found a real solution, and was executing it with the tools his position gave him. That the solution required draining someone else's valley was a cost he accounted for differently than the people living in it. He and the valley's farmers were reading the same hydrology and arriving at incompatible conclusions about whose survival it served.

Los Angeles voters approved a $1.5 million bond issue to purchase Owens Valley lands and water rights. The vote was 10,693 for and 754 against. Better than ten to one. The rival Examiner disclosed that Times publisher Harrison Gray Otis held financial interests in San Fernando Valley land that would be served by the aqueduct. Even the Examiner eventually supported the project. The Times warned before the vote that a negative voter would be "an enemy of the city."

The valley had no organized voice in the Los Angeles bond campaign. The people whose water was at stake lived 240 miles away, in a string of small towns most Los Angeles voters had never visited. Nobody was required to ask them. Valley protest became visible in federal and statewide venues by 1906 and 1907, when the San Francisco Call reported Owens Valley charges against the Reclamation Service. But by then the bond had passed. A democratic majority had ratified the extraction. What had been Eaton's private land deals became the city's official project, backed by voter mandate and municipal credit. Reversal now required overturning the expressed will of a city that outnumbered the valley by orders of magnitude. The ratchet clicked.

The Withdrawal

On May 25, 1907, Theodore Roosevelt issued Proclamation 756, setting apart the Inyo National Forest. The designation covered public lands along the Owens River. A local valley history argues that Forest Service chief Gifford Pinchot, with Roosevelt's encouragement, made former reclamation lands, "mostly treeless," part of the national forest rather than returning them to the public domain for homesteading.

The practical effect was surgical. The valley could not grow into the land the federal government had withdrawn. It could not homestead. It could not build the agricultural base that might have given it the economic weight to resist Los Angeles. The federal government, having declined to build the valley an irrigation project, now ensured the valley could not build one itself. The land that might have been the valley's leverage became land the valley could not touch.

Each of the three decisions had its own logic. Eaton was acquiring water rights at market prices for a city that needed water. Los Angeles voters were securing their future. Roosevelt was applying conservation principles his administration believed in. The valley was on the losing side of every calculation, and by 1907 the range of possible futures had narrowed to one.

The Resistance That Was Already Foreclosed

The aqueduct opened on November 5, 1913. Mulholland's dedication was five words:

The valley fought back in the 1920s. On May 21, 1924, about forty men traveled from Bishop and dynamited the aqueduct's concrete canal north of Lone Pine at 1:30 in the morning. On November 16, 1924, Bishop banker Mark Watterson led farmers and ranchers who seized the Alabama Gates spillway. An estimated 500 to 800 valley residents picnicked and camped at the gates for four days. The occupation ended November 20. The aqueduct was repaired. The water kept flowing south.

The reasons the resistance failed were structural. Los Angeles already controlled the key reservoir site, fifty miles of river frontage, and the voter-backed financing to buy more. The federal government had withdrawn public land from settlement. The Watterson brothers' attempt to build an alternative financial base through the Owens Valley Irrigation District collapsed when the courts found their bond sale invalid due to W.W. Watterson's direct interest as a district director. By the end of 1926, Los Angeles controlled 90 percent of the valley's land and water.

The dynamite and the occupation are the parts of the story that get remembered, because they look like a fight. But the three decisions of 1903–1907 had already removed every institutional lever the valley might have pulled: the federal irrigation alternative, the public-land development option, the possibility of matching Los Angeles's financial and political weight. What remained was physical resistance against a legal and economic structure that had already been built around the valley like a canal around a river. The water was going where the structure sent it.

Mulholland and Eaton, meanwhile, had their own falling out. Eaton set a $1 million price for the Long Valley land Los Angeles needed for a storage reservoir. Mulholland refused to pay it. The first aqueduct operated without the Long Valley reservoir Mulholland had wanted. Los Angeles eventually bought Eaton's ranch after the Watterson bank failure for less than the $500,000 Mulholland had previously offered. The winning side had its own competing rationalities, and they produced their own casualties.

What the Landscape Became

Owens Lake was dry by the mid-1920s. The dissolved minerals and salts that had been accumulating for millennia in a lake with no outlet crystallized into an alkali salt crust across the exposed lakebed. Where water had reflected the Sierra, there was now a hundred square miles of white and tan playa, crusted and cracked, baking under the same sun that used to glint off the surface. When the wind crossed the dry bed, the crust broke apart and went airborne.

The EPA would later determine that Owens Lake emitted approximately 300,000 tons of PM10 particulate matter per year, including 30 tons of arsenic and 9 tons of cadmium. The largest single source of PM10 pollution in the United States. The dust storms plagued the valley for more than half a century before enforcement began. The people who remained in the valley breathed what the lake had become.

LADWP signed a 1998 memorandum of agreement accepting responsibility to reduce dust emissions. The EPA approved a control plan in 1999 using water, gravel, and vegetation on parts of the dry lakebed. LADWP's current dust mitigation program covers 48.6 square miles of exposed playa and has reduced emissions by 99.4 percent. It has cost Los Angeles ratepayers more than $2.6 billion.

Two and a half billion dollars to keep a dead lakebed wet enough that the wind doesn't poison the people who still live in the valley the city emptied.

Water goes onto the playa and evaporates, which is what it did when the lake was alive, except now it's called dust control and it's budgeted. The system that was destroyed has to be partially simulated, at enormous cost, by the entity that destroyed it, because the consequences of the destruction turned out to be unlivable. Nobody maintained the lake. Nobody was chartered to protect it. It was just the thing that happened when the water arrived and had nowhere else to go. It was so reliable it was never named, and that made it the most dangerous thing to lose.

The Owens Valley story is sometimes told as a crime. Chinatown made it noir. But the record supports something harder to metabolize than villainy. Eaton was buying water rights at market prices. Mulholland was solving an engineering problem with an engineering solution that worked exactly as designed. Los Angeles voters were choosing their own future by a margin of ten to one. Roosevelt was applying conservation principles his administration believed in. Lippincott's dual role was a conflict of interest, not a conspiracy. The Reclamation Service chose the larger population over the smaller one, which is what utilitarian governance does. Every actor was reading the system accurately. They were making different calculations about who the system served, and the calculations that won were the ones backed by more people, more money, more institutional weight.

Every decision was defensible. The cascade they produced killed a lake, emptied a valley, and created the worst particulate dust source in the country. The people who lived in the path of that cascade understood what was happening to them. They dynamited the aqueduct. They occupied the spillway. They lost, because by the time they fought, the structure that determined the outcome was already built around them. Each click of the ratchet had been accompanied by a reasonable explanation.

Owens Lake is still there. It is 48.6 square miles of managed playa, shallow-flooded and graveled and tilled, maintained by the city that drained it at a cost that now exceeds what the entire aqueduct originally cost to build. The wind still blows. The dust control holds, mostly. The valley is still mostly owned by Los Angeles. The water still flows south. The system that replaced the old one requires constant intervention to prevent the consequences of its own existence, and nobody who built it planned for that, because the lake was not infrastructure. It was just a lake. It was just the thing at the bottom of the basin, doing what a terminal lake does. Collecting. Evaporating. Persisting. Until it wasn't.