Irrigation efficiency for who?

I recently met Bruce Lankford, an experienced irrigation expert, at a small conference in London. He was annoyed that policy makers misunderstand the basic facts of irrigation, and especially in how they encourage “efficient” practices that may not be. They assume, for example, that drip irrigation is “modern and efficient” whereas flood row irrigation is “traditional and inefficient.”

He sent me his 2012 paper: “Fictions, fractions, factorials and fractures; on the framing of irrigation efficiency” [pdf].

While reading it, I realized that the debate he’s engaging is not “policy relevant” to economists like me (totally in line with his brilliant Table 1).

Let’s break this down:

Irrigation specialists, managers and the farmers who use irrigation systems really need to pay attention to how much water goes where.

• Managers need to make sure the system delivers water fairly (with respect to rights) and evenly (with respect to operational performance).
• Specialists care because they want to have consistent measures of performance, which are needed if one seeks to improve performance or redesign systems.
• Farmers care about efficiency to the extent that it allows them to get “more crop per drop.”

None of these concerns need be relevant to policy-makers and economists, but they often are.

Why?

Policy makers/politicians love free-lunches, and “efficiency improvements” allow them to give more water from the same volume. With water scarcity and food demand rising, who would not want to grow more food with the same quantity of water? But that goal is vulnerable to accounting errors and tricks, which often means that “win win” systems (on paper) are “win lose” or “lose lose” in reality.

What’s going wrong? Hayek would say that information is too diffuse and unstructured to be aggregated and (effectively) used by centralized decision-makers. And I would agree, which is why I have — for years — favored the simple, but crude policy of allocating irrigation rights with the assumption of 0% efficiency, i.e., 100% consumption of the water.*

This assumption means that “saved” water is a gain to the farmer with rights, another farmer, or the environment. It also means that “efficiency gains” do not come at the expense to non-irrigators.

For example:

• Ten farmers each have rights to 5 units of water in a watershed with 100 units of water.
• If efficiency is 0%, then all 50 units are used, and 50 units are left for the environment.
• If efficiency is 60%, then each farmer “uses” 3 units and 2 units are “lost.” Overall consumption of 30 units leaves 70 units for the environment.

A win for ecosystems? Typically no, since a policy maker would say “Wait! We can allocate those 20 units of saved water to 7 farmers (rounding from 6.67), with the assumption that they also use 3 and lose 2, so now farmers are using 50 (it’s really 51), and 50 still goes to the environment! Win win!”

But now say that farmers get “more efficient,” and reduce losses from 2 to 1 unit per farmer. Now 17 farmers use 68 units. If they reduce “waste” to zero, then they use 85 units! The environment suffers, even though rights are exactly the same.

This process is not just made up — it’s the default, since policy makers like pleasing farmers, and the environment is a terrible lobbyist.

That’s why I support 100% consumption/0% efficiency when allocating rights.* In that case 50 units are allocated and “used” on paper, while the environment gets 50 units on paper. If the farmers “waste” 2 units each, then 20 units ALSO go to the environment, so my conservative accounting provides better protection to the flows that sustain ecosystems — and us.

Farms, as businesses, are profiting from their private use of water that is, fundamentally, a resource owned by citizens. That means that they should not operate if that damages ecosystems (qua public goods). Although that is often the case, it should not continue in a climate changing world, due to the rising value of the ecosystems we still have.

Oh, and my “one unit, all consumed” (OUAC!) preference also helps with water markets and trading. If rights are somehow defined in terms of (non-zero) efficiency, then trading or transferring rights is difficult without complex re-balancing to reflect the efficiency of the buyer. With OUAC, those changes are interesting to farmers, but policy people can ignore them.**

And let me be clear: efficiency, losses, flows and other measures of what water goes where is totally important and useful to farmers and irrigation managers. They need to open and close gates; they need to maintain head flows; and they need to make sure crops get the right volume at the right time, but all of those details — pace Hayek — need not be aggregated to policy makers or into water rights.

My one-handed conclusion is that water rights should be simple to understand and compatible with other water uses, without creating risks and crises if and when practices, technologies or use-location changes.

---

*Read a few of these:

• From farms to cities (Apr 2008) — how to do it right
• Craig Wilson is wrong about ag water use and so is Peter Gleick (Jan 2011)
• The irrigation efficiency conundrum (Bredehoeft Feb 2011)
• Water rights: diversion or consumption? (2012)
• Chris Perry on water rights (Perry May 2018)

**A change of point of use (or time of use) does have impacts, even with OUAC rights. I still support those adjustments, which are calculated with formulas, for example in the Murray-Darling Basin (Australia).