Updating my Colorado River reservoir storage spreadsheet today to make a graph for a friend was a pretty discouraging exercise. Lake Mead and Lake Powell, the system’s two massive reservoirs on which 9 states and a gazillion people and farm acres depend, are at their lowest combined level since 1967, when they were filling Powell for the first time:

Colorado River storage

I style myself as the optimist in the room. Staring dimly into the future, I think I can see what solutions to the West’s water problems might look like. My optimism comes from things like this, which appears on the agenda of tomorrow’s meeting of the Imperial Irrigation District board of directors. It’s an information item about an Arizona plan to conserve close to 100,000 acre feet of water this year, primarily by fallowing agricultural land in central Arizona:

This is water that will sit in Lake Mead, unused. It’s the kind of step that encourages me, but the details of what’s needed to carry it out help explain why Lake Mead is dropping faster than my optimism can keep it propped up.

It’s encouraging because it shows flexibility in the water use system. To slow or stop Lake Mead’s fall, we need to use less water by doing things like this. Central Arizona’s got an agricultural buffer to work with, imported Colorado River water pumped uphill via the Central Arizona Project. This agreement calls on some of that buffer.

But the friction here is discouraging. This is not as simple as just saving water. Dustin Garrick’s fascinating new book, Water Allocation in Rivers Under Pressure, goes a long way toward articulating what I’ve only been seeing dimly as I watch this process. Garrick’s work takes a deep look at the “transaction costs” of moving water around and, more importantly, in making the institutional changes (the changes in governance structures and rules) needed to change the way we move water around. (The “Coase” of the headline is economist Ronald Coase, whose work on transaction costs is incredibly important here if you wanna go full wonk on this stuff.) One of the key takeaways from Garrick’s book is that it’s not enough to point to how the rules are all messed up and Lake Mead is dropping as a result. We’ve got to really understand why the rules are hard to change. Those are transaction costs.

We’ve got a mechanism here that’s relatively straightforward on the water conservation side – fallow land in Arizona, leave water in Lake Mead. But the bureaucratic side is a nightmare of transaction costs. Load up the full document linked above and scroll down, and you’ll see that in order for Arizona to get credit for this arrangement under the Colorado River’s formal water accounting system, the agreement needs 16 signatures.

Sixteen signatures is a lot of transaction costs.