If you had asked me to guess whether public or private utilities did better at water conservation, I would have without hesitation guessed that public utilities did better.

So here’s a fascinating result from Manny Teodoro and Youlang Zhang of Texas A&M, looking at data from the recent California drought:

[O]n average, communities served by private utilities adopted more stringent conservation regulations than those served by public utilities; and so … private utilities were significantly more likely than their public counterparts to meet the state’s conservation standards; and … private utilities on average conserved more water than public utilities. Somewhat counterintuitively, then, private, profit-driven firms were more effective than were local government agencies in achieving the state’s conservation goals.

The working paper, Privatization as Political Decoupling: Water Conservation and the 2014-2017 California Drought, is here. Their argument is that, for a public utility governed by directors facing reelection, rate increases and other conservation constraints on consumers pose political risk. Private utilities, on the other hand, offload that political risk to a state regulation commission, which governs their rate setting.

We argue that, where policy goals can be achieved through regulation of private firms, private provision of public services allows governments to separate public policies from their political costs. By shifting production or service provision from the public to the private sector, governments can achieve policy goals through regulation, while shifting the accompanying political risks to the private sector, where they are less acutely felt. The result is a political decoupling that allows governments to achieve policy goals while insulating officials from their political costs. One implication is that, where financial decoupling exists, regulated private firms are more likely to comply with environmental regulations than are government agencies, because the latter bear electoral costs that the former do not.

This result makes sense, though it is not at all consistent with my naive expectation. But it is consistent with a similar finding from Megan Mullin and Meghan Rubado, who found that water agencies with directly elected boards of directors in Texas responded more slowly to drought.