Two farmers plant alfalfa. One doesn't worry about his water supply; he can use as much as he wants on his crop. The other also can use all the water he wants, but he has to pay a tax on it. Who is likely to make more money?
If your answer is the first farmer, you are in good company. For decades, economists could find no monetary benefit to managing access to groundwater. But, as researchers at Binghamton have proven, you and all those economists are wrong.
"We know that groundwater is a problem around the world," environmental economist Neha Khanna says. "Groundwater is being depleted. You can talk to anybody on the street and they'll say, 'I know that, and we need to start thinking about it.' Strangely, though, economists have always believed that if we put in some sort of system to manage water, it wouldn't lead to much gain in terms of overall welfare. We have pretty much ignored the issue."
In the 1980s, several prominent economists analyzed the possibilities. They found such marginal gains from managing access to groundwater that it became an established result among economists: There's no reason to implement water-management systems. They even gave it a name, the Gisser and Sanchez Effect, after the authors of a seminal 1980 paper.
A different approach
Models in the economics literature have treated the physics of water flow in a simplistic way, notes Khanna's colleague Andreas Pape, assistant professor of economics at Binghamton. The traditional method—the "bathtub model"—supposes that when you draw water from an aquifer, the water level drops evenly thoughout. But a bathtub is a poor stand-in for an aquifer, he says. For one thing, the depth of the water is irregular because water moves gradually through the ground rather than instantly like a giant underground pool. Changes in the type of rock or soil surrounding the water also affect how quickly the water flows through the ground.
"If you take a straw and suck some water from a glass, you can see the level of the water dropping evenly," Khanna says. "But that doesn't happen in an aquifer. The water is flowing through the materials in the soil: the sand, the clay and so on. It's more like when you take a sip of a smoothie through a straw. You can sometimes see a little depression right around the straw. And that's what happens with groundwater around a well."
Hydrological models, on the other hand, do a much better job of mapping and predicting water flow in aquifers. Such models are of vital importance to states that rely heavily on agriculture, and these states often employ hydrologists who study water resources in great detail. Their models, however, don't generally take economic concerns into account. For example, Khanna says, these calculations wouldn't be concerned with how "expensive" water is; that is, whether it is easily accessible or must be pumped from a great depth.
An interdisciplinary team at Binghamton set out to build an economic model with a more sophisticated view of the physics of water flow, one that could take advantage of the latest geographic information systems (GIS) data. Economists Khanna and Pape were joined by their doctoral student, Todd Guilfoos, as well as hydrologist Karen Salvage.
What they came up with combines an optimization model—that's the traditional economics piece—and a simulation model, which is where the hydrogeology comes in. Tying it all together was Pape's expertise in agent-based modeling, a technique that's becoming more mainstream even though its use in economics is still fairly new.
After developing confidence in the model during trials with hypothetical scenarios, they were ready for a real-life challenge. Guilfoos settled on California's Central Valley Aquifer, where Bakersfield is the major population center. Kern County, Calif., had GIS maps that included detail about which crops were being grown. Some crops use more water than others—alfalfa needs more water than carrots, for example—Guilfoos says, and that information was used as a proxy for demand for water in the model. They also included information on well location, long-term precipitation that recharges the aquifer and much more. With help from Kevin Heard, assistant director of Binghamton's GIS Campus Core Facility, the economists imported these GIS maps into their model.
"Twenty years ago, maybe even 10 years ago, we didn't have the computing power to do this," Khanna says. But these new tools provide a way to reexamine issues in economics, even ones that seem to have been settled definitively.
One scenario in the Binghamton model supposes there's no management: Farmers take as much water as they want. The other, Pape says, involves taxing the water extracted at each well to curb overuse of the aquifer. "Then," he says, "we compare the long-term profits of the farmers under the two scenarios and measure the difference."
Previous work showed very small gains from water-management plans. The new model shows savings of several magnitudes higher when the hydrogeology is taken into consideration. It predicts increased profits for Kern County farms involved in a water-management system—over the course of perhaps 80 years. "That's how far into the future we're looking," Guilfoos says, "but they actually start to see profits much, much earlier than that, within a decade or so."
Farmers understand at one level that they have an impact on each other, he says, but because aquifers are so large, the individual farmers don't necessarily see that their use has much of an effect on water levels. The model shows that heavy users would receive more benefits from a management system, he says.
And when the economists talk about "benefits," they mean actual dollars and cents, not just a feeling of moral superiority: "We found that there is, in fact, a lot to be gained in terms of economic welfare from managing water," Khanna says.
How is that possible? Pape puts it this way: "Let's say there are two neighboring farmers. Each is trying to decide, 'Should I withdraw another gallon of water today?' Suppose that if he does, he adds $1 to his future pumping costs and $1 to the pumping costs of his neighbor. So the public cost—the total social cost—of the pumping is $2, but the private cost—the cost facing just him—is $1. Left to his own devices, the farmer will consider the cost to be $1 and impose the extra cost on his neighbor. Since pumping seems cheap to him ($1 instead of $2), he pumps more than he would otherwise. However, his neighbor is making the same decision! As a result, they both impose extra costs on each other.
Both neighbors would be better off if they chose to pump less, that is, if they recognized that the cost of pumping is $2, not simply $1. The policy remedy in this case, therefore, is to assign a tax of $1. Then both farmers will be deterred from overdrawing water, making them both better off."
The tax is essentially a tool to bring about the conservation of water, Guilfoos notes. "This conservation of water decreases the cost of extracting water in the future, and people don't have an incentive to save water for the future on their own," he says. "Conserving water reduces farmers' profits now but increases the farmers' total profit in the long run. And we demonstrate that conserving groundwater can be significant to long-term profits, which is new to economics."
As the group begins to look at other aquifers, Guilfoos says, they are seeing some variations. "The significance of the gains depends on the aquifer," he notes. "We did find significant gains in Kern County. We found small overall gains in another aquifer in Pecos, Texas. Not all aquifers are going to have large gains; it depends on the dynamics of demand, well placement and how fast the water moves."
Guilfoos has presented the Kern County research at several conferences, and the team recently submitted a paper to Environmental and Resource Economics, a flagship European journal, for review. Guilfoos is focusing on new data from Kansas now. Through the Kansas Water Office, he obtained detailed maps on well location, soil types and crops. This has enabled him to develop detailed economic models for the Kansas section of the Ogallala aquifer, the largest in the United States.
"Now we can say it's important to manage water," Khanna says. "But the next question is, 'How do we manage that water?' As an economist, the ultimate driver for me is a desire to affect policy."
As the team looks to policy recommendations, they'll try to assess what can be done reasonably, given the current economic and political environment. Which policy should be pursued? Should there be a price for water? Should states consider managing well locations? How do you get the stakeholders to talk to one another, especially when the boundaries of an aquifer don't necessarily line up with county or state lines?
Guilfoos notes that the implementation need not be mandated from the top down. There are lots of ways to manage water resources, he says, from the local level up to the federal level. He and his colleagues do see an opportunity to influence policy, likely starting with the Kansas project, as there's already the political will there to make changes.
Once there's an American example, Khanna envisions putting the research into practice in arid regions of countries such as China, India and Spain. She says, quite simply: "We need to look at this collectively."
Explore further: Regulators order pipeline testing, other steps after spill