Understanding progressive & regressive water pricing
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:
- Uniform, where customers pay the same price for every unit of water that they consume;
- Inclining block, which charge higher prices as volume increases; and
- 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.
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.
U.S. water utilities are shifting costs to low-volume customers—good for revenue stability, but bad for affordability
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?
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.
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 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.
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
An update on what low-income U.S. households must pay for essential service
About a year ago I also published the results of a national study of affordability using these new metrics using data from a nationally-representative sample of utilities in 2017. This year, I’ve been working with Texas A&M graduate student Robin Saywitz to update that study for 2019 with fresh data and an expanded sample of utilities. A working paper details the full methodology and results for the 2019 update; this post reports the main findings.
Water and sewer affordability remain at the forefront of discussions among utility policymakers, managers, and regulators as communities across the country face rapidly rising capital, maintenance, and replacement costs. Since good policy requires good measurement, I’ve spent a lot of time in recent years evangelizing for a new, double-barreled measurement approach published early in 2018: the Affordability Ratio and Hours at Minimum Wage. These metrics are designed to capture the sacrifices and trade-offs that low-income households face when paying for these essential services.
Sample & data
There’s no central repository for water and sewer service rates in the United States, so valid depictions of affordability requires gathering data directly from utilities. We drew our sample from the EPA’s Safe Drinking Water Information System (SDWIS), which contains basic system information for the country’s nearly 50,000 water systems. To get a representative picture of the nation’s affordability we stratified the sample and then randomly selected systems. Our analysis then adjusted mathematically for the sampling procedure to make sure we get representative results.
We collected rates data during the summer of 2019 with an active survey—that is, we gathered rates information directly from utilities rather than relying on responses to questionnaires. The final dataset included full water and sewer rates data for 399 utilities—an increase of 70 systems compared with the first study.*
Basic water & sewer share of disposable income
The first main metric is the Affordability Ratio at the 20th income percentile (AR20), which estimates basic water and sewer prices as a share of disposable income, which for present purposes is total income minus other essential living costs (including taxes, housing, food, health care, and home energy), estimated using Consumer Expenditure Survey data.
In 2017 AR20 averaged 9.7 in the United States; we found that average AR20 is up sharply since 2017, with the national average now at 12.4. In substantive terms, that means that in the average U.S. utility, basic water and sewer service prices for a four-person household consume about 12.4 percent of disposable income for a household at the 20th income percentile. Here are the distributions of AR20 for 2017 and 2019:
As the figure shows, the overall distribution has shifted markedly since 2017, especially at the high end. AR20 is less than ten for about 60% of systems, but we found a troubling growth in very high values of AR20.
Basic water & sewer in hours at minimum wage
The other way I like to measure affordability is to convert basic monthly water/sewer prices into Hours’ Labor at Minimum Wage (HM). It’s not a precise way to measure affordability, but it’s intuitively meaningful.
The overall HM distribution also shifted up overall, with an average HM of 10.1 in 2019. In other words, the average monthly price of basic water and sewer service for a four-person household in the United States requires about 10.1 hours of labor at minimum wage. Average HM was 9.5 in 2017, so we see an increase in 2019, but the change in average HM is not as dramatic as the change in AR20. Here’s affordability in the United States measured in terms of local minimum wage, again with 2017 and 2019 side-by-side:
So what’s going on with affordability?
Trying to infer trends from just two time periods is tough, but some important patterns are evident in the 2019 update. The most striking result overall is the sharp increase in AR20, which reflects increases in water and sewer prices, increases in essential expenditures, and stagnant or even declining 20th percentile incomes.
The more modest increase in average HM is consistent with nominal inflation, but in terms of minimum wage labor the increase of +36 minutes is noteworthy as minimum wages have not increased in many parts of the country.
We’re still unpacking what it all means, and I’ll have more to say about what the 2019 data show. For now, these figures offer another snapshot of water and sewer affordability in the United States to help frame discussion over improvements and structural reforms to the American water sector.
*399 might not seem like a very large sample for a country with 50,000 water systems, but the sample is much larger than most studies of water rates, it’s representative, and we’re highly confident in the validity of the data because we gathered the data ourselves. It’s also 70 more systems than we sampled in 2017.
© 2019 Manny P. Teodoro