From Rates

New Normals?

​Lessons from California ​water conservation​, 2019

​Reservoirs aren't supposed to look like this.

Tough water times may be back in California. After the Golden State suffered through a historically severe drought from 2012-2017, ​pleasantly wet weather in 2018-2019 refilled reservoirs and replenished mountain snowpacks. But the state’s drought monitor shows that the past few months have trended drier, and water managers are worried that the state might be slipping back toward another drought.

One fortunate legacy of California’s recent struggles with drought is the California Water Board’s investment in data collection and dissemination. Researchers are learning important lessons about water management from that wealth of data on water consumption and conservation. Complete data for 2019 are now available, and I’ve just had a first look.

Three things emerge from my initial cut at the 2019 data that merit mention against the ominous backdrop of a looming drought.

1. ​A way of life

California’s overall urban water use remains down significantly from the pre-drought days. The state government​ ​established emergency conservation ​rules during the peak of the drought, cutting statewide urban water consumption by nearly 25%. But Governor Jerry Brown and other leaders also vowed to make water conservation “a way of life” that would extend beyond the emergency.

Californians seem to have taken the goal to heart, at least in aggregate. Though water conservation is not at the zealous peak we saw in 2015-2016, it remains strong:

Data: California Water Resources Control Board

Falling conservation in 2018 led to some hand-wringing, but conservation rebounded in 2019 to a statewide aggregate 18.1% relative to the state’s official 2013 baseline. That’s pretty remarkable and suggests that much of the reduction in California’s urban water demand is more or less permanent. Absent severe drought or monsoon-like precipitation, California’s overall water conservation will likely remain in the mid-to-high teens for the foreseeable future. That’s a public policy success story.

2. Persistent public-private disparity

One of the most fascinating findings that emerged from my earlier analysis of the California drought with Youlang Zhang and David Switzer was that the state’s private, investor-owned utilities conserved significantly more water than local government utilities during the drought.* We linked that difference in drought response to the institutions that govern water finance. We also found that the public-private conservation gap persisted even after the drought ended, with the greatest disparities during the summer when water demands are at their highest. Back in 2018 your humble blogger forecasted that that “public and private conservation will converge in the spring and diverge again in the summer and autumn.” And indeed, the pattern held: there was essentially no difference between public and private water utility conservation during the winter months, but during the 2018 May-September peak season, California’s investor-owned systems saved more water than their local government counterparts.

Did the pattern persist in 2019? In short, yes:

Data: California Water Resources Control Board

As you can see, public and private utility conservation move in pretty close parallel. During the non-peak period (January-April and October-December), there is essentially no difference in average conservation by ownership. But a public-private gap emerged again during the May-September peak period. In 2018 the difference was 2.3%; in 2019 the gap was smaller—investor-owned utilities saved an average of 1.5% more water than local governments. To give that percentage some context, think of it this way: if public utilities saved at the same rate as private utilities in 2019, the difference would have been about 15 billion gallons, roughly equivalent to the City of Long Beach’s annual water consumption.

The public-private disparity in summertime conservation now also appears to be a way of life in California. The difference is almost certainly related to pricing and revenue. Decoupling über alles.

3. Indoor outpacing outdoor?

​Finally, the five years of data now in front of us show how California’s urban water conservation relates to overall seasonal demand. In 2015, when severe outdoor irrigation restrictions were in effect for much of California, 70% of the state’s conservation came from reduced demand in the peak season. Since then, the share of peak season consumption has fallen to about half:

Data: California Water Resources Control Board

In 2019 non-peak conservation crept up to 52% of overall savings. It’s hard to infer exactly what’s driving this pattern from aggregated monthly data, but the picture suggests that the most persistent water savings seen in California over the past five years has come from indoor efficiency. Average water consumption in 2019 was about 98 gallons per person per day (gpcd), but some cities are far lower. For example, water demand in conservative San Francisco shows hardly any seasonal peak at all, and averages just 42 gpcd—a level approaching the theoretical minimum to sustain developed world conditions.

While there are probably still plenty of savings to be had from indoor efficiency, meeting the challenge of another potential drought year—and driving further sustainable conservation—will require tackling peak demands through pricing and regulation. Get ready for another year of water cops and rate revolts.

* Water geeks Interested readers can see the full study in Policy Studies Journal for the details.

​© 2020 Manny Teodoro

Introducing the Amy Poehler Index

Understanding progressive & regressive water pricing

It's irrigation season, everybody!

By Antonio & Manny Teodoro

How do utilities distribute the costs of drinking water systems to their customers in their rate structures?

The answer is surprisingly complicated, and water utility pricing is often weird. There are lots of other wrinkles and variations, but the vast majority of utilities use one of three basic rate structures:

  1. Uniform, where customers pay the same price for every unit of water that they consume;
  2. Inclining block, which charge higher prices as volume increases; and
  3. Declining block, which charge lower prices as volume increases.

It’s easy to see that these different rate structures distribute costs differently, but how much differently isn’t immediately obvious. How do a utility’s rates apply to low-volume customers compared with high-volume customers?


The answer is important because it carries ​significant implications for affordability and conservation. It also speaks to risk tolerance and questions about fairness. Water is an unusual consumer good because its use is very different at different volumes. For residential customers, low volumes are mostly used for basic needs like drinking, cooking, cleaning and sanitation. Higher volumes are usually for more discretionary uses like lawn irrigation and car washing.

Studies of water rate structures typically put them into the three main categories (uniform, inclining, and declining), which is fine, but can mask some important variation within the inclining and declining blocks. Consider four imaginary rate structures:

A and B are both inclining block rates, but A is more progressive than B because prices increase sharply for A as volume increases. Similarly, C and D are both declining block structures, but C is more regressive than D because ​C discounts higher volumes much more than D.

David Switzer developed a way to measure water rate progressivity​ to reflect that variation, and published a paper last year that uses regression slopes to measure relative progressivity. It’s a creative, rigorous, and smart methodology, but it’s pretty sophisticated and not the easiest approach for communicating with the general public.

In search of a valid but more intuitive way to communicate the idea of progressivity, we struck upon the idea of comparing average unit costs of water at relatively conservative and very high benchmark volumes. What would be appropriate comparative volumes? And how could we frame the measurement in an engaging way?

Enter Amy Poehler

​Amy Poehler was a particularly profligate water customer. In the summer of 2015, while drought ravaged the Golden State, the Parks & Recreation ​star’s Beverly Hills home used 85,000 gallons a month. Meanwhile, a family of four that is fairly conservative with water uses something like 6,000 gallons per month for drinking, cooking, and sanitation.* In other words, Poehler’s home consumed in about two days enough water to comfortably supply a family of four for a month.

​Poehler's house. Maybe she's got a dialysis clinic in the basement?

Shaming celebrities for bad environmental behavior is now something of a ritual in America, and it’s not clear whether exposing excess actually helps. But Poehler’s water consumption provides a convenient benchmark for excess.

The Amy Poehler Index

So to measure progressivity we calculate the total monthly water and sewer bill—including both fixed and volumetric charges—for a customer at 6,000 gallons (a conservative family) and at 85,000 gallons (Amy Poehler’s family), then divide that price by each customer’s total volume. These are average unit costs. The ratio of the two unit prices is the Amy Poehler Index (API). A value of 1.0 means that Amy Poehler and the conservative family pay exactly the same unit price for water. Values less than 1.0 indicate regressive rates (Amy Poehler pays less than a conservative family), and values greater than 1.0 indicate progressive rates (Amy Poehler pays more than a conservative family).

Let’s look at how this works for a couple of large U.S. city water systems under their 2019 rates:

In 2019 Tampa’s fixed monthly charge for water was just $1.50, with no fixed charge at all for sewer. Under Tampa’s inclined five-block rate structure, Amy Poehler would pay $617.82 monthly, while our conservative family would pay just $19.29. On a unit cost basis, those prices equal $7.27 and $3.22 per thousand gallons, respectively. The resulting API is a progressive 2.26.

​Meanwhile, in 2019 Philadelphia charged its customers a fixed $5.12 for water and $7.04 for sewer each month. The City of Brotherly Love then applied declining block water rates that would have charged Amy Poehler $455.69 monthly, and the conservative household $41.10—more than twice the Tampa bill for the same volume. The resulting unit costs turn out to be $5.36 for Amy Poehler and $6.85 for our conservative family, for a regressive API of .78.

The National Progressivity Picture

We used data from the Teodoro & Saywitz 2019 affordability update to calculate API for a nationally representative sample of 399 U.S. water and sewer systems. Average combined water and sewer rates were slightly regressive at .88, but ranged widely from .07 in Anchorage, AK to 3.81 in Phoenix, AZ.

API isn’t as precise as Switzer’s progressivity coefficient, but in our national dataset the two metrics correlate pretty well (ρ=.71). More importantly, the API offers an easy way to understand and improve the ways that communities distribute costs through their rate structures. That seems like the sort of thing Leslie Knope would probably dig.

*That’s about 50 gallons per person per day (gpcd) for a four-person household. 50 gpcd is an indoor efficiency standards for California and Texas.

© 2019 Antonio Teodoro & Manny Teodoro

Squeeze Play

U.S. water utilities are shifting costs to low-volume customers—good for revenue stability, but bad for affordability

Luke Skywalker & friends escaped this crunch thanks to everyone’s favorite droid, but R2D2 isn’t going to help utilities escape their financial problems

Rising water and sewer prices linked to increasing capital and operating costs are driving affordability concerns across the United States, and with good reason. Studies of water rates typically measure prices at benchmark volumes that are meant to reflect “average” residential customers.* But for purposes of low-income affordability, how a utility structures its prices across levels of demand is as important as what it charges an average customer or how much total revenue it pulls in.

Over the past year I’ve been working with Texas A&M graduate student Robin Saywitz to analyze 2019 water and sewer rates data.† Among other things, we’re comparing our recent dataset with similar data from 2017. Although it’s difficult to infer trends from just two time periods, we’re seeing a troubling pattern in U.S. water and sewer rates: not only are prices increasing overall, average prices are rising much faster for low volumes than for high volumes.

That’s very bad news for affordability. Why are utilities squeezing their low-volume customers with higher prices?

Colliding trends

The answer starts with two broad water sector trends that have converged to drive water prices to their present point. First, long-deferred capital maintenance and upgrade costs are finally coming due, and long-deferred water and sewer revenue needs are rising accordingly. Utilities need more money to pay for pipes and people. Emerging challenges like lead service line replacement and new contaminants like PFAS only make things more expensive.

At the same time, average urban water demands have been falling steadily over the past twenty years. Back in the 1990s when I first got into the water business it was an article of faith that long-term water demand increased with economic and demographic growth, and long-term supply adequacy was a paramount concern in many parts of the U.S. The water sector responded with a widespread push for conservation. Thanks to organizations like the Alliance for Water Efficiency, we’ve seen an astonishing decline in average water demand—especially for essential indoor use. For the first time, America has seen sustained urban growth with steady or even declining overall water consumption. That’s an extraordinary accomplishment, and it’s rightfully celebrated.

But the combination of rising costs with declining average demand creates a revenue problem for water utilities. Declining total demand means that the average price of water must increase steeply in order to generate needed revenue.

Perils of progressive pricing

For years, utilities have been pushing for progressive water rate structures to distribute costs equitably and to encourage conservation. Indeed, progressive pricing is part of why we’ve seen declining demand. As I’ve observed before, water service is unusual in that its use varies considerably at different levels of demand. For residential customers, low volumes reflect essential uses like drinking, cooking, cleaning, and sanitation. Higher volumes are typically associated with discretionary uses like car washing and outdoor lawn irrigation. So progressive rate structures that charge relatively low prices for low water use, steeply higher marginal prices for high volume use, and volumetric sewer charges generally result in better affordability. What’s more, good rate design helps affordability without the transaction costs, administrative burdens, and social stigma that come with means-tested assistance programs.

But progressive rate structures raise utilities’ revenue risk. Revenues from volumetric charges fluctuate vary seasonally and can skyrocket or plummet depending on the weather. A utility doesn’t sell much high-priced, high-volume water if it rains all summer and nobody waters their lawn. Even worse, sales can fall sharply during drought emergencies when customers conserve water. That can leave the utility in tough financial shape, because the utility’s capital and operating costs are mostly fixed. Progressive pricing can put the squeeze on utilities’ revenue needs.

The squeeze

So utilities are, in turn, putting the squeeze on their most conservative customers with more regressive pricing.

The first gallon price of water and sewer service is a useful touchstone to understand the real impact of rate structure changes. 

The squeeze bunt puts more pressure on the hitter than any other play in baseball. Utility financial managers can relate.

​The first gallon price is the price a customer pays for using any water at all: any fixed charges plus the price of the first unit of water or sewer service. For example, if there is a $20 monthly fixed charge for water service and the first thousand gallons of water is $2.00, then the first gallon price for water service is $22.00. Here are the weighted average prices of water and sewer service in 2017 and 2019 at one gallon, 6,000 gallons, 12,000 gallons, and 20,000 gallons:

​Unsurprisingly, the first gallon price increased from $35.80 to $40.89 over the two-year period, and average prices went up at each volume level. If prices were simply going up across-the-board, we’d see roughly equal increases in prices at every volume. But the 2019 data show that price increases were uneven in percentage terms:

At 20,000 gallons monthly, average prices went up by 8%, but the first gallon price increased by more than 14%. As prices have increased, low-volume customers have on average borne a much larger share of utilities’ rising revenue burdens than their more profligate neighbors.

Managing risk

The financial challenges associated with equitable, affordable, progressive pricing are real: utilities can’t survive without revenue, and falling or fluctuating demand creates real risks for sustainable utility management. But there are better ways to manage risk than squeezing the most conservative customers.

A rate structure that provides basic volume allowances at low fixed prices with steeply inclined prices at higher volumes is one good option. As I’ve observed before, consolidation can help maintain progressive pricing because larger customer bases can withstand revenue shocks more easily than small systems. Utilities should also use larger cash reserves to stabilize revenues across seasons and years—and governments should keep their hands off those reserves! More creative approaches could include regional water revenue banks or development of a secondary market for utility revenue risk.

*A lot of studies claim to measure “average bills,” but are really measuring bills at specific volumes that are assumed to reflect an average customer. Studying true average bills across large numbers of utilities is hard because there’s no reliable source of data on average consumption across utilities.

An initial working paper reports the full methodology and descriptive findings in detail.

© 2019 Manny P. Teodoro