9 Feb 2012

Water rights: diversion or consumption?

One of the first complications we run into when discussing water rights is how to quantify them.*

Consider an example of an irrigation canal that carries 100 units of water shared among 10 farmers. If each farmer gets an equal share towards diverting that water, then each farmer can take 10 units from the canal.** But these diversions are not the same as consumption. Irrigation with 10 units of water does not mean that the crops will consume 10 units. Consumption varies with irrigation technology. Drip irrigation may use 95 percent of the water (with the rest evaporating or sinking into the soil); flood irrigation may only result in 20 percent consumptive use -- most of the water flows off the land and back into the canal. These return or tailwater flows mean that 100 units of diversion will not result in 100 units of use; excess tailwaters flow downstream. Farmers who do not like this "waste" tend to allocate water in terms of consumption, so that a farmer with 10 units of consumptive use (and 66 percent irrigation efficiency) may divert 15 units of water to use 10 units. Farmers allocating 100 units of consumptive water will not leave any "waste" to their neighbors (or the environment).

Those are the facts. The interesting question, now, is how best to encourage efficiency among those farmers.***

We can create an incentive to use less water by raising the price per unit of water diverted to the farmer, allowing farmers with a right to 10 units of consumption to sell some of that consumption, or by imposing an assumed "efficiency parameter" on a farmer, i.e., assuming that a farmer with 10 units of consumption rights only needs 12 units of diversion to get 10 units of consumption.

Each of these solutions (individually or together) suffers from an important flaw -- the need to track water consumption. Such tracking is important when a farmer claims he is only using eight units of water and wants to sell the other two or only pay for eight units. It's also a problem when the farmer uses more water than assumed by the efficiency parameter (11 of 12 units, for example).

It thus seems easier to speak in terms of a farmer's water diversion -- something that's easier to measure -- and then ignore the amount of water the farmer uses. Thus, we can have 10 farmers taking 10 units each and ignore their return flows, assuming that anything that makes it downstream is a bonus.**** Such a system would be easier to administer. Efficiency would rise if farmers found ways to increase consumption for each diverted unit (reducing their payments for diversions) or sold their "unused" tailwater to others. Sales of diversion-based water rights outside of the area would tend to cause problems downstream by removing all return flows; such an event could be avoided by quantifying sales/export in terms of consumption, but that number would require measurements of consumption that was not tracked in the past.

These are just some thoughts on how to measure rights for markets or administrative purposes.

Please add your own thoughts, corrections and questions.

* It's worth noting that the word "right" can create emotions out of proportion to their function. It may be better to call them entitlements or licenses, to make it easier to retire "paper water" that will never be allocated with wet water or to facilitate trades by reducing the emotional burden of selling one's "right" (see my discussion in my paper on All-in-Auctions).

** We ignore "head water" flows in the canal that keep the canal wet and make it possible to carry 10 units to each farmer.

*** We are ignoring environmental flows, which are either taken care of separately from the farmers' decisions or are augmented by purchases from farmers who use less.

**** Mike Young, while discussing this point, indicated that downstream users (farmers or environmentalists) would count on these flows as rights. They would object to a decrease in flows that might result from greater efficiency (more use) by the 10 farmers.