Using Auctions to Share Scarce Water

bighornplateau1 (via flickr)
An aqueduct snakes through Southern California. As water becomes more scarce, we will need new systems for sharing water equitably and efficiently.

In the summer of 2009, we were treated to a bicoastal spectacle of conflicts over water in the western and southeastern United States. Farmers in California were on television, complaining about a “congressionally induced” drought; the mayor of Atlanta was arguing that Florida’s wetlands were taking water his citizens needed to survive. In 2011, the water shortages shifted to the Midwest and northern China. And these global water crises are here to stay if we continue to ignore the signs of increasing water scarcity and continue to manage water as if it is as abundant now as it was in the past.

One of the bumper stickers above my desk says “Nature makes a drought; man makes a shortage.” We may not be able to do anything about drought, but we can do something to manage the resulting scarcity. Economics—the so-called science of scarcity—has a number of tools that can prevent scarcity from turning into shortage. The challenge is to find the right way to put them to use.

An “all-in-auction” mechanism—a simple way to ration scarce water among claimants—represents a radical departure from the good old days, but it also recognizes and reconciles the new reality: we no longer have abundant water to quench every thirst.

In most countries, water is owned by the People, but the “usufruct right” to use water is distributed to individuals. Some areas use formal rules or informal institutions to ration water among claimants within the community. This is the case in urban Southern California, for example, where 26 cities and municipal water districts share water imported from stressed areas (the Colorado River and Sacramento River Delta). In other areas, politicians and bureaucrats distribute usufruct rights to users—as is the case in the “suddenly dry” southeast, where Georgia, Florida, and Tennessee find themselves squabbling over rights to Lake Lanier, the exact position of state borders, and the relative importance of Atlanta subdivisions and Florida river flows.

The end of abundant water means that usufruct rights are increasingly scarce and valuable. Existing owners face challenges from new claimants who claim they can put water—a social good that provides simultaneous benefits to different groups—to higher and better uses. Additional pressure comes from existing users—farmers, urban water managers, or environmentalists, for example—who claim that they should get more water, even if that means that their neighbors would get less.

Such scarcity naturally leads to shortage if competing demands are not rationed to meet supply. Regulatory or political rationing can be distorted by inaccurate rules or biased lobbying; price rationing is transparent but tends to favor uses that generate cash over uses that generate noncash benefits (water for the environment or subsistence farming, for example). Doing nothing (continuing to use water for flood irrigation, for example) is unacceptable. Water belongs to all of us, and we want to make sure that it is used efficiently.

An all-in-auction can fairly and efficiently reallocate water while addressing existing property rights, environmental demands, changes in supply, and the claims of new users. The all-in-auction does all of this by selling all water to the highest bidders (facilitating reallocation among uses), allocating sales proceeds to existing owners (acknowledging property rights), and allowing existing owners to “buy from themselves” at no cost (allowing any owner to opt out and maintain historical water use).

All-in-auctions are best suited for use in small geographic areas and should be repeated on a short interval. Local auctions make it easier to reduce accidental impacts on neighbors, maintain hydrological integrity (uses are close to the sources), minimize transportation costs (infrastructure and pumping energy are expensive), and maximize local participation. A short time interval means that the quantity of available water is known (instead of being based on a seasonal average, for example), but the interval also needs to be long enough to ensure reliability (farmers want water for an irrigation cycle or growing season; urban water managers want to meet demand using stored and purchased water).


Olaf Gradin
The 2007 drought across the southeast caused Atlanta’s main source of water, Lake Lanier, to drop to record levels.

All-in-auctions are for bulk water. They are not for people worried about water for their shower. They are also, primarily, for water transported in rivers and canals, but they can be used for allocating water from an aquifer, reservoir, or desalination plant.

Take, for example, a group of 50 farmers who share water from an irrigation district canal. The supply of water varies with time and weather. It may also be reduced by regulatory or logistical deductions for environmental uses and system losses, respectively. (An irrigation system needs a minimum amount of water in the trunk canals to allow additional water to be delivered to smaller canals.)

Let’s say that there are 1,000 units of water available. Scarcity implies that these 50 farmers want more than 1,000 units of water. If there is plenty of water, then there is no scarcity and no need to do anything but deliver as much water as each of the farmers desires (while covering the cost of infrastructure). In a small irrigation district (with five farmers, for example), it may not make sense to run an auction. The five farmers can bargain for water over a few mugs of coffee more quickly than they could arrange this auction.

Rights may be allocated by seniority, historical use, acreage, and so on, but let’s assume that these farmers share the water equally. Since there are 50 farmers, each one gets an allocation of 20 units of water, an amount that pleases no one, since everyone wants more water.

For simplicity, we will limit participation in this auction to the farmers, which means that there are no environmental, legal, or political impacts from their decision to reallocate their water rights within their group. That said, it would be easy to allow outsiders to buy water for use in a city, on other farms, or for the environment.

So, now we are ready to have an all-in-auction for those 1,000 units of water with 50 farmers, each of whom have rights to 20 units of water. The key to this auction design is that all 1,000 units are available for sale (“all-in”), and all units are sold to the highest bidders. (More exactly, the 20 units of available water are rented to others; the usufruct rights remain with the original owner.) No farmer can hold onto his 20 units of water; they are for sale. He must bid if he wants water. These two characteristics mean that the auction market is as liquid as possible (pun intended) and that water goes to those who are willing to pay the most for each unit.

But what about property rights? Each farmer receives the revenue from the sale of his 20 units of water, which means that he is paid for his water. This feature also makes it easy for a farmer to bid as much as he wants for 20 units of water, because the money he pays will be exactly the same as the money he receives. He does not need to worry about losing his water to a sneaky last-minute bid. This “no-regrets” auction only ends if nobody bids for five minutes.

It is important to stop here and clarify these features. We want to put all of the water up for sale to make sure that the auction reallocates as much water as possible. At the same time, we want to make sure that farmers who want to keep their water (those who would not offer it for sale in normal circumstances) are able to “buy it from themselves” at little or no cost (the cost of their time plus some administrative cost for running the auction).


Lance Cheung/USDA Photo
Arrowhead Farms owner Akram Mohammad (left) used a micro-irrigation system to save his 719-acre pecan farm from the recent droughts in Texas. An all-in auction system promotes conservative water use, while making sure that water rights are allocated fairly, thus protecting the claims of farmers like Mohammad.

So we are going to have 50 farmers making many bids (more than 1,000 bids because water is scarce), and the highest 1,000 bids will win the auction. For accounting convenience, all of these winning bids will pay the same price, P, which is set equal to the 1,001st-highest bid. This design is the same as we see with eBay, where the winner pays a price based on the second-highest bid. In our case, it is the 1,001st-highest bid because there are 1,000 units for sale.

Thus there will be net sellers who buy less than 20 units of water and earn P for each unit they do not buy. If P is $10, then Bob will earn $80 if he only buys 12 units. By the same accounting, there will be net buyers who buy more than 20 units. If Chuck wins 25 units, then he will pay $250 into the auction and receive $200 (for his property of 20 units), leaving him with a net payment of $50 and 25 units of water. Finally, we have Andy, an indifferent farmer who wins 20 units, taking $200 out of one pocket (as a buyer) and putting $200 into the other pocket (as an owner). Note that Andy can bid as much as he wants ($300 per unit!) because he is going to earn as much money as he spends. High bidding ensures that Andy gets his units but it also makes Andy think a little about whether he might want to sell a few units of water for $10 each. Then Andy, Bob, and Chuck go back to their farms to put their water to work.

The all-in-auction makes it easier to reallocate water to different users while recognizing existing property rights and allowing for environmental and weather-related adjustments to supply. These auctions allow current holders to opt out, rent their property for cash, or buy additional water at competitive prices. They minimize political and bureaucratic interference and can be implemented anywhere, by any water user group—farmers, urban water managers, industrial users, and others. These auctions for water enjoy several advantages over auctions for carbon permits because water is a local good.1 A water auction sets a price in one place without affecting prices or uses elsewhere. These auctions also take existing property rights into account; carbon markets have not established these rights. Auctions also make it easier to tax the increasing value of usufruct rights to water—rights that, in many cases, were given away long ago.

Auctions are well suited to allocating fluctuating water supplies among changing uses and users. That’s because each auction begins by quantifying the amount of water available to the current set of owners—a number that may be higher or lower than the last auction—and allocates that water to those currently interested in using the water. The mix of buyers and sellers may change from auction to auction; the mechanism makes it as easy as possible to reshuffle water among them. Developing countries with unstable water supplies and evolving uses would have much to gain from this system, as it makes it easy to recognize property rights and adjust to changes in available volumes, while also ensuring that the water goes to those who are willing to pay the most for it. Auctions also promote transparency in property rights, water pricing, and final allocations—characteristics that facilitate the development of robust institutions, undermine corruption, and advance the welfare of citizens.

All-in-auctions can be implemented right now by existing water-distributing organizations under their current legal and logistical structures, to maximize the efficiency with which we use our increasingly scarce and valuable water.