In arid states where water is running low, irrigation is close to the whole ballgame.
In arid states where water is running low, irrigation is close to the whole ballgame.
O ne of the top reasons Americans cite for opposing data centers is their excessive water use. Yet the notion that data centers are guzzling the nation’s reservoirs is, to put it softly, a fiction.
Based on the most recent figures, data centers consume 17.4 billion gallons of water each year. That sounds like an enormous amount, until you realize the nation consumes tens of trillions of gallons. Even including data centers’ indirect water consumption via the electricity generation needed to power them, they account for less than 1 percent of the country’s total water consumption. Data centers use less water than the nation’s golf courses or California’s almond industry.
Yes, data centers’ water demands are expected to grow dramatically over the coming years, but it will remain a tiny percentage of overall use, and they are becoming more water-efficient through recycling techniques. More importantly, data centers must go through the same process of acquiring water as any other business or development, either by buying water rights from existing holders or by applying for a new withdrawal permit, depending on the location.
If not to data centers, Americans might wonder where all their water is going. Western residents are inundated with stories about the Colorado River running dry, forcing deep cuts to withdrawals. That river is a vital water source for seven states: Arizona, California, Colorado, Nevada, New Mexico, Utah, and Wyoming. In 2023, the governor of Arizona moved to halt new housing construction around Phoenix for fear of a groundwater shortage. Communities across the country are implementing voluntary or mandatory drought restrictions on households to conserve water.
Many people assume that most of the country’s fresh water goes into the public supply available to homes and businesses, since that is their primary interaction with water. In truth, public supply accounted for 12 percent of water withdrawals in 2015, the most recent year for which full accounting is available. All the water from utilities that flowed into bathrooms, kitchens, and swimming pools, sprayed onto lawns and gardens — all those millions of gallons make up just 12 percent of total supply.
The lion’s share of withdrawals — 41.3 percent — is for thermoelectric power. Electric plants heat large volumes of water into steam to spin their turbines, then use additional water to condense the steam back into liquid and repeat the cycle. Because the water they use is recycled, power plants ultimately return the vast majority of it to the source. Thus, the sector considered the largest water user in America is not the largest consumer.
That honor belongs, far and away, to agriculture. From 2010 to 2020, crop irrigation consumed more than ten times as much water as public supply and power generation combined.
In Western states, water use is even more skewed toward irrigating the desert. For the eleven states traditionally counted as the American West, irrigation’s share of water withdrawals ranges from 59 percent in Washington to 96 percent in Montana. (Agriculture’s share of water consumption is likely higher.) For comparison, public supply is responsible for no more than 20 percent of water withdrawals in any Western state, and as little as 1–2 percent in Wyoming, Montana, and Idaho. As for the particularly depleted Colorado River, three-quarters of withdrawals are for irrigation. Alfalfa alone, a type of hay grown primarily to feed cattle, accounts for roughly half of the river’s water use.
If dry states have loads of water to use for relatively low-value crops like hay and wheat, which wetter states produce in far larger quantities, why are residential users always the first to bear the brunt of droughts? The answer is that water rights in the American West, dating back to the 19th century, are not allocated for efficient use. Rights are established under the doctrine of prior appropriation, meaning that the first person to use a water source for a beneficial purpose won the right to use that water in perpetuity. All subsequent water users had to get in line.
When the West was populated by agrarian settlers, prior appropriation worked just fine. After almost two centuries of urban population growth, however, water has become scarce. Farmers have clung tightly to their land and associated water rights through the generations, leaving the thirstiest newcomers — cities and suburbs — to fight over what remains.
To address these worsening shortages, water must be treated economically as the scarce resource it now is — no longer as an unlimited public good. That requires putting a price on it and letting a new market determine the most efficient use based on who is willing to pay.
Such water markets can be implemented without rescinding farmers’ longstanding usage rights. In fact, they could make existing water rights more valuable. By allowing holders to trade freely within the same general basin, farms could sell or rent their water rights to cities and other users when prices are high enough. The least valuable uses would be given up first, while crops with greater returns would stay fed. Once faced with an opportunity cost, farms of all kinds would be incentivized to adopt more efficient irrigation practices and profit from the water they don’t use. By contrast, the current legal principle of “use it or lose it” encourages overconsumption.
Another drying country shows the way. In Australia’s Murray-Darling River Basin, water rights are classified as permanent entitlements unbundled from land, which afford owners an annual percentage of available water. Both those yearly allocations and the underlying entitlements can be exchanged across several interconnected hydrological markets, overseen by a basin authority. Entitlements are collectively valued in the tens of billions of dollars, and rights are constantly traded among farmers, governments, industry, investors, and even conservation groups.
Ignore the water demagoguery. Halting the construction of data centers won’t replenish the Colorado River or free up critical supplies for Western towns and cities. The only thing that can is a shift away from resource-inefficient crop irrigation, which is best facilitated by fluid markets.