One thousand Arkansas rice farmers have just about pumped away all their ground water. Naturally, they are looking to taxpayers for relief.

Specifically, they are clamoring for the U.S. Army Corps of Engineers, the federal agency responsible for America's waterways, to spend $200 million on a new system of reservoirs, canals, and pumping stations to divert water to their farms from the nearby White River. The state is supposed to toss in an additional $119 million, for a total taxpayer gift of $300,000 per farmer. This is on top of the considerable crop subsidies the farmers already receive: They sell their rice for $3.10 per bushel even though the market price is only $1.40.

Approving the White River project would only compound more than a century of bad federal water policy. Resource economist Delworth Gardiner, a professor emeritus at the Brigham Young University, has calculated that the total cost to society of a typical federal irrigation project is $400 per acre-foot of water (an acre-foot is the amount of water it takes to cover one acre to a depth of one foot). The market value of the water ranges from $50 to $100 per acre-foot, but farmers usually pay the Bureau of Reclamation about $20 to $30. If such irrigation projects were evaluated on a true cost-benefit basis, says Gardiner, "there would be no new federal water projects at all."

The Arkansas rice farmers' situation is a microcosm of a problem affecting much of the world. A new report from the International Food Policy Research Institute, "World Water and Food to 2025," addresses the global freshwater crisis. "Unless we change policies and priorities, in twenty years, there won't be enough water for cities, households, the environment, or growing food," Mark Rosegrant, lead author of the report and a senior research fellow at the IFPRI, warned in a press release. "Water is not like oil. There is no substitute. If we continue to take it for granted, much of the earth is going to run short of water or food—or both."

Experts estimate that more than 41,000 cubic kilometers (km3) of renewable water are available annually. That's about three times the size of Lake Superior. But that water is not distributed equally. For example, Brazil and Canada together account for 25 percent of the world's renewable freshwater. Farmers are by far the largest users of the world's freshwater. The problem is that critical farming regions such as northeast China and northwest India are running out of freshwater to grow food.

In the IFPRI report's business-as-usual scenario, total water withdrawals will increase by 22 percent, from 3,906 km3 of water to 4,772 km3 by 2025. With appropriate policy changes, however, total withdrawals would increase only to 4,042 km3, which would leave more water available for environmental purposes.

To avert the crisis, water has to be properly priced, but such an approach would encounter strong resistance. Both Gardiner and the IFPRI report point out that once subsidized water projects are built, the value of the water allocated to farmers, even if it's not a formally recognized property right, is incorporated into the value of their farms. After the land is sold, the people for whom the subsidized infrastructure was built are no longer the beneficiaries. The new owners have paid premium prices for irrigated land, not land that could revert to desert at the whim of a federal administrator. One could argue that a farmer who lives by the subsidy should die from its withdrawal. Whatever the merits of that view, implementing it is not politically feasible.

So what to do? "Key to inducing higher water efficiency gains in all sectors is introducing market (or market-style) incentives into water use decisionmaking," argues the IFPRI report. "Incentive prices for water could have a major impact on water withdrawals and consumptive use in irrigation and urban water uses, thus freeing water for environmental use."

One ingenious end run around the political problems posed by subsidized water allocation is a "charge-subsidy" scheme that would involve allotting a base property right to a certain amount of water, taking into account its historical use, to individual farmers or groups of farmers. If they used more than their base amount, they would pay market prices for the additional water. If they used less, they could sell the water they saved at market prices to other users, such as cities and industries. This arrangement would strongly encourage water use efficiency and establish water markets between farmers, urban dwellers, and industrial users.

In the case of federal water projects, farmers would receive their initial allocation of water at, say, $20 per acre-foot. If they needed more, they'd pay $100 per acre-foot. If they saved water, they could sell it to city dwellers for $100 per acre-foot. In fact, it might be more profitable for some farmers to stop farming and sell all their water to other users. Such a system would not be perfect, but it would be better than the mess we find ourselves in now.

Water markets face opposition from some "progressives." The current issue of Mother Jones, for example, features a package of articles attacking the privatization of water resources. The articles focus mostly on city water systems, but one article asks, "Should water, a basic necessity for human survival, be controlled by private interests?" In another article, Indian ecofeminist Vandana Shiva complains, "There's a global water crisis, and at the same time water is being increasingly commodified."

Fortunately, some environmental groups, such as the World Wildlife Federation, are coming to see the importance of markets for conserving water and protecting ecosystems "Only water markets can create the incentives to reduce waste and send water to its most productive and necessary uses," writes Capital Research Center economist David Riggs. "By treating water as a tradable commodity, by including water in market processes, people across the world and the natural environment will be better served."

Which brings us back to the Arkansas rice farmers. "Why should we subsidize a pump that will sell subsidized water to grow a subsidized crop?" a local opponent of the White River project asked The New York Times. That's exactly the right question. We should not go down the dead-end road we've already traveled with so many other federal water projects.

Central Arkansas typically receives 50 inches of rain annually. That's more than enough to raise many other crops, such as corn, wheat, and cotton. It's not as though these farmers have a heaven-mandated right to grow rice. If they can't make a profit, they should be allowed to go bankrupt like any other small business.