This paper [PDF] joins the Santa Barbara study on my must-read list.
Kenney et al. make the following points:
- They assume that customers respond to the average price of water that they see on their most-recent water bill. They make this assumption because their academic colleagues could not calculate (!) the marginal price of water on sample bills they were asked to examine, and because customers have no idea of what their current use is (more on that later).
- Low water users have the lowest price elasticity (-0.34) for total water use. High water users have the highest elasticity (-0.75). When use restrictions are in place, these elasticities change, rising to -0.46 for low users and falling to -0.24 for high users. (Restrictions lower demand by 6.5 and 14 percent, respectively.) The implications are that water managers should target high-water users with restrictions in drought (short-run) but higher prices in other times (long-run). Low-water users can be targeted with higher prices, but that's probably not a good idea for two reasons: They cannot make much of an impact in overall water demand (low-water users are, by definition, not using much). Second, higher water bills are more likely to penalize poor people, who tend to use less water (smaller yards, etc.).
- Further, average elasticities nearly double in drought conditions, rising from -0.56 to -1.11. This is probably due to a greater degree of customer awareness of water in the media, etc.
- Kenney et al. find that the presence of increasing block rates (not prices) will lower demand by 5 percent. That's probably because people are leery of ending up in a higher block.
- Finally (and paradoxically), they find that customers with "smart meters" actually use MORE water. Why is that? Because they are able to get closer to the maximum consumption in their block without surpassing it. This finding does not damn smart meters. Instead, it indicates that customers will be more responsive to rates and blocks when they know how much water they are using.