EPA-670/1-74-001
April 1974
Environmental Health Effects Research Series
PRICING FOR WATER SUPPLY
ITS IMPACT ON
SYSTEMS MANAGEMENT
N$
#&> s*fc
I
55
\
$
SSEZ
'i'x r*
^ PRO^°
13
LU
CD
^r
/
National Environmental Research Center
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
EPA-670/1-74-001
April 1974
PRICING FOR WATER SUPPLY;
ITS IMPACT ON SYSTEMS MANAGEMENT
by
Robert M. Clark
Haynes C. Gpddard
Program Element No. 1CA046
NATIONAL ENVIRONMENTAL RESEARCH CENTER
OFFICE OF RESEARCH AND DEVELOPMENT
U. S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
NQTICE
This report has been reviewed by the
National Environmental Research Center,
Cincinnati, and approved for publication.
Mention of trade name? or commercial
product^ does not constitute endorse-
ment or recommendation for use.
ii
-------
FOREWORD
Man and his environment must be protected from the adverse
effects of pesticides, radiation, noise and other forms of
pollution, and the unwise management of solid waste. Efforts
to protect the environment require a focus that recognizes
the interplay between the components of our physical envir-
onment—air, water, and land. The National Environmental
Research Centers provide this multidisciplinary focus through
programs engaged in
• studies on the effects of environmental contaminants on
man and the biosphere, and
• a search for ways to prevent contamination and to recycle
valuable resources.
This report examines the costs associated with supplying
drinking water to consumers and their reaction in terms
of increased or decreased consumption based on the prices
which they must pay. It also explores the relationship of
drinking water quality to cost and suggests the possibility
of using the pricing mechanism as a means of inducing waste-
water reuse thereby increasing the life of the available
water resource.
A. W. Breidenbach, Ph.D.
Director
National Environmental
Research Center, Cincinnati
• it
111
-------
ACKNOWLEDGEMENT
The authors would like to acknowledge
the assistance of Arthur F. Hammonds of the
Water Supply Research Laboratory, NERC-
Cincinnati, who provided computer support
for the analysis presented here.
IV
-------
PRICING FOR WATER SUPPLY;
ITS IMPACT ON SYSTEMS MANAGEMENT
INTRODUCTION
Problems related to water supply have become increasingly
important in recent years. In the past, water has been so
abundant that it was available in almost unlimited quantities.
But this is no longer the case in most parts of the United
States today. Water has become a resource that is relatively
scarce. And the land, labor, and capital resources needed to
convey water to places of useful application and to collect
and treat wastewater are also scarce. To help deal with these
problems, the National Water Commission was established by an
Act of Congress and approved by the President on September 26,
1968. Its duties include reviewing national water resource
problems, making projections for future needs, finding alter-
native ways of meeting these needs, considering economic and
social consequences of water resource development, and advising
on specific water resource matters.
If a resource is scarce, society must apportion its use so
that it yields the maximum beneficial return. The Commission,
in its draft report-"--^ has concluded that the limited supply
of usable water should be rationed among its most important
and productive uses, where it will have maximum utility for
society. The pricing mechanism is a powerful and remarkably
effective way to do this.
The rationale behind the pricing mechanism is based on the
assumption that the more units consumed of any commodity,
the less valuable is the last unit consumed. Thus, as the
amount of water used by a consumer increases, the value of
the last drop he uses decreases. The phenomenon is referred
to by economists as "diminishing marginal utility."
When a price tag is attached to water, each consumer will
continue to use more and more until its value to him just
equals the price he is charged. When the price of water
-exceeds its value to the consumer, he will stop using it.
The value of other goods and services that the consumer can
command with his limited financial reserves begins to exceed
the value of the extra units of water.
-------
But those whose use of additional water continues to yield
a return or utility in excess of the cost of the additional
water will use more. Thus, water use will be shifted where
it is most productive in terms of aggregate utility to society.
Congress has directed the National Water Commission to con-
sider conservation and more efficient use of existing supplies
as an alternative way of meeting future water requirements.H
Though other means might be employed to motivate better use of
existing and future supplies of water (elaborate rationing
mechanisms based on input-output analysis, for example),
nothing is as simple, comprehensive, and effective as the
pricing mechanism.
This paper examines some of the work done by other investigators
who have studied the interactions between pricing and water
consumption. One objective is to verify and confirm some of
the contentions made by these investigators that relate price
and consumption of water, and that relate the cost of water
supply to various physical factors in the system. Another aim
is to examine the effect of pricing on design parameters for
water supply systems and its possible effect on the design
of wastewater reuse systems. Data used in this analysis was
taken from the Community Water Supply Survey conducted by the
Bureau of Water Hygiene in the summer of 1968. But before
this data is discussed, a brief examination of the history
and current status of water supply systems will be useful.
HISTORY OF WATER SUPPLY
Historically, water supply and wastewater disposal have their
roots in antiquity, where the design, construction, and man-
agement of public water supplies and wastewater disposal systems
were allied to the growth of capital cities and religious or
trade centers. Developed as installations of considerable
magnitude and complexity, their remnants still stand as monu-
ments to sound yet daring feats of early engineering. A
notable example among the great structures of ancient times
are the aqueducts and sewers of Rome and her colonies.4
The sanitary control of water supplies and wastewater is of
recent origin and is associated with the growth of cities
during the industrial revolution of the 19th century. The
scientific discoveries and engineering inventions of the late
18th and early 19th centuries paved the way for the creation
of centralized industries to which people flocked for employ-
ment. As a result, the community facilities of the mushrooming
industrial cities became overtaxed, and the need for the abundant
distribution of safe water and for the effective disposal of
-------
human excrement and other wastes became imperative* Water
was often drawn front polluted rivers or from shallow wells
in crowded sections of the community and distributed in covirt-
yards by standpipfes 6n intermittent days. The fatigue asso-
ciated with fetching water Was so great that the inhabitants
only used it when absolutely nedessary or for cooking; they
rarely used much for washing clothes or personal hygiene.
Much of the water used for domestic purposes was of question-
able quality, and waterborne disease outbreaks became common.
To combat the growing disease problem, the science of sanitary
engineering was born, and With it, the design and construction
of the precursors of modern water supply systems.
CURRENT STATUS
Today, water is supplied to Municipalities for drinking and
culinary uses, bathing arid washing, heating arid air-cOnditioning
systems, watering df lawns and gardens, recreational use in
swimming and wading pdpls, numerous and Varied industrial
processes, and protection df life and prc-pefty against fire.
To provide for these varying uses, Which average about 700
gallons per dwelling unit per day/ the supply df: water must
be satisfactory in quality and adequate in quantity; readily
available to the user, relatively che'ap, and easily disposed
of after it has served its purposesi The engineering systems
that serve these multipurpose objectives atfe the water-works
or water-supply system, and the wastewater works, or waste-
water Collection and disposal System.
The water supply system collects water ftfom its natural
sources, purifies it if necessary1 > arid delivers it to the
consumer The wasteWater system collects seWage or used water
from the community-^abdut 70 per dent of the water Supplied,
together with varying amounts of ground and surface water
that enters the system. Surfade1 rUnoff resulting front rain-
storms and melting snow arid ice is either collected by a,
system of drains that also carries away residential and in-
dustrial wastewater (combined sewerage) dr collects the runoff
separately in an independent or Separate/system df storm drains
(separate sewerage), the collected sewag^ or wastetaater is1
often discharged, after suitable treatment* i^ntp a natural
drainage channel, a receiving body df water, or, on occasion,
onto land. Often the saitte body of water serves both as a
source of water and as & recipient of sewage and storm drainage.
It is this multiple u^e of watdr that establishes the most
important reason for water sanitation; Pollution df water by
waste sources makes the wett6r unsightly, and malodorous £nd has
the potential of adding disease-producing organisms to water
suppliesi
-------
There are many interactions between municipal water supplies
and wastewater treatment systems. These interdependencies
grow more important as a region becomes more urbanized. The
connecting link between the water supply and wastewater dis-
posal systems is the plumbing system or system of water supply
and wastewater collection within dwellings, commercial estab-
lishments, and industries.
Water supply systems are generally composed of (1) collection
works, (2) purification works, where needed, and (3) transpor-
tation and distribution systems. The collection works either
tap a source of water that is adequate in quantity to satisfy
present and reasonable future demand on a continuous basis,
or they convert an intermittent source into a continuous supply
by storing surplus water for use during periods of low flows.
If the water is not of satisfactory quality at the point of
collection, purification works treat it: Polluted water is
disinfected; esthetically unattractive or unpalatable water
is treated to make it attractive and palatable; water con-
taining iron or manganese is subjected to deferrization or
demanganization; corrosive water is stabilized chemically;
and excessively hard water is softened. The transportation
and distribution works convey the collected and treated water
to the community, where it is distributed to consumers. The
amount of water delivered is often measured so that a charge
can be made for its use.
Municipal supplies may be drawn from a single source or from
a number of different sources. The water from multiple sources
is ordinarily mixed before distribution.
BASIC PRINCIPLES OF PRICING
In a competitive economy, the* forces of demand and supply,
acting through the impersonal and automatic market mechanism,
determine prices. These serve as proper guides to potential
users and producers in their consumption, production, and
investment decisions, since prices are based on the proper
balance of a consumer's valuation of a product against its
costs. Hence, prices automatically vary with changes in
value and cost, and resources are directed to their "best" uses
In some sectors of the economy, such as with public utilities,
prices are not determined by the impersonal forces of the
market. Instead, they are set by the pricing policies of
utility managers. To retain the attributes of efficiency
that are associated with competitive prices, these utility
rates should be varied with changes in the demand and cost
conditions facing the utilities. Even When a utility is
producing a single product in a physical sense, there will
-------
be variations in supply and demand conditions that are asso-
ciated with the quantity of the product consumed during
different times and at different locations. Therefore, for
purposes of setting prices that reflect incremental, not
historical costs, it is necessary to distinguish between
products consumed at different locations and times. Although
products might be physically similar, they are economically
distinct and must be treated separately if they are to be
efficiently priced.
One of the most important dimensions of supply and demand is
temporal and will be used for illustrative purposes in this
section. There are other dimensions that can be discussed,
but the basic principles that will be established here apply
to those dimensions as well.3
When the quantity of a good demanded at any given price varies
over time, a "peak-load problem" exists. The solution requires
one to choose the "right" capacity and the "right" price struc-
ture, a choice that determines the level of output in each
subperiod of the production cycle. In--competitive markets,
the peak-load problem is solved by prices that fluctuate over
time to reflect changes in value and cost. Theater and motion
picture houses charge more for performances during periods of
peak demand (evenings and week-ends) relative to periods of
off-peak demand (matinees and weekdays). Resort hotels have
higher rates during peak seasons. Rental fees for recreational
goods exhibit a similar pattern, and barber shops frequently
charge more on Saturdays than on other days.
Even when nominal prices remain constant, real prices may
vary, since the quality of service may be altered during
peak-demand periods. For example, grocery store prices remain
constant during the store hours, but queuing- and congestion
increase the cost of shopping during peak hours. The open-
market solutions to the peak-load problem vary widely. Some
competitors provide no service at off-peak, and some provide
the same service as at peak but at a lower price. Others
provide better service at off-peak, but charge peak prices.
Possibly all three alternatives may be provided, each by a
different group of competitors. All of these solutions are
the results of responsive and flexible pricing in which, in-
cremental values are balanced against incremental costs.
Thus real prices are lower in times of excess capacity and
correspondingly higher in times of shortage.
The conventional solution to the peak-load problem by urban
water utilities, which do not operate by the rules of the
open market, is basically different from the foregoing examples
The problem is handled solely as one of supply management.
-------
That is, peak-demand days are usually taken as given—they
are not related to price—and supplies are adjusted to meet
these "requirements," regardless of the incremental values
and costs associated with doing so- Not much imagination is
needed to realize that resort hotels following this planning
procedure and not utilizing demand management strategies would
soon find themselves facing bankruptcy. Demand management, or
the alteration of demands through responsive pricing, is not
considered as being within the range of a water manager's choice.
Prices typically remain constant from season to season, and so
resources are used inefficiently, and inequities are imposed
on the utility's consumers.
Uniform water rates mask significant differences in the mar-
ginal costs of serving customers during different periods.
If summer users want more water, additional capacity must be
provided; whereas only the incremental costs of operating the
existing system must be incurred for additional use. Summer
water, in an economic sense, is significantly different from
winter water: Summer water is high-cost water. By not varying
water rates to reflect these cost differences, investments are
larger than economically justified. That is, peak demands that
are used to guide water utility investments are not appropriately
restrained by prices that reflect the peak-load marginal costs.
The general public, as well as most utility managers, perceives
uniform rates as being equitable, since the same price per unit
of water is charged to all customers, regardless of when the
water is consumed. Because of cost differences, however, uni-
form prices mask a subtle type of price discrimination. Winter
consumers subsidize summer consumers, because the winter con-
sumer is absorbing a portion of the capacity costs created by
the summer consumer's demands. In many cases, this means that
the inner-city consumer who has a small yard to irrigate and
a relatively small summer peak demand subsidizes the more
affluent suburban dweller who has a larger yard and a greater
summer peak demand.
The efficient allocation of resources that would be obtained
in an open-market competitive economy can be approximated by
water utilities if water managers broadened their range of
choice to include demand management. In short, to solve the
peak-load problem in an efficient manner, the following prin-
ciples must be applied: (1) if capacity is not fully utilized,
the price should reflect operating costs with no contribution
to capacity costs; and (2) if demand exceeds capacity at this
price, the price should be adjusted upward to restrain demand
to the capacity level. Therefore, if the same type of capacity
serves all users, capacity charges should be levied only when
capacity is fully utilized, so that these peak users bear the
responsibility for defraying capacity costs. When the capacity
-------
charges exceed the incremental costs of capacity, investments
in capacity are justified; whereas the reverse case would indi-
cate that existing capacity is excessive. (This investment
rule is for constant-cost conditions and would have to be
modified slightly for decreasing or increasing costs.3)
PRICING OF WATER SUPPLIES
Current Practice
Present water pricing is far from the ideal desired for an
effective pricing system. Pricing by public agencies is
typically based on revenue considerations, since the primary
aim is a balanced budget and customer satisfaction. Pricing
to provide efficient checks and balances on resource alloca-
tion is not given very high priority and is commonly below
the cost of amortizing and operating a water supply system.
Funds are usually sought elsewhere to make up the difference.
Self-supplied users, who account for about 80 percent of all
water withdrawals (54 percent by industry, mostly for cooling;
23 percent for self-supplied irrigators; and 1 percent for
rural supplies other than irrigationlO), usually pay no price.
However, they do generally bear the full costs of their own
diversion and delivery systems and thus have, in effect, an
internal price equal to costs of obtaining supply. Users of
water, public or private, are now typically awarded a right
to divert and use water free of charge; they, therefore, need
pay no attention to values that some other use of the water
might yield. Furthermore, only infrequently do means exist
for the sale of water rights to bring about a reallocation
of its uses. As a result, withdrawals from the natural water
system are not always allocated to the uses that can yield
the highest return.
The ability to apply refined pricing systems to the total
supply is limited, since pricing of self-supplied water to
reflect opportunity costs of water would require legislation
that not only adopts a policy in favor of pricing, but also
establishes entities to levy and collect charges.! Neverthe-
less, existing municipal and industrial supplies controlled
by a water service agency can be subjected to improved pricing
policies.
Influence of Pricing on Municipal Water Use
*
There is strong evidence that metering and pricing have sub-
stantial impacts on water use. Introduction of metering, for
example, reduced water use by 36 percent in Boulder, Colorado.2
Reductions ranging from 20 to 50 percent have been achieved in
-------
13
other areas by metering, which accomplishes two things.
First, users are made aware of the extent of their water
use. Second, water charges are, in effect, changed from a
flat-rate system of pricing to rates based on incremental
use. Both the information and the financial incentive are
important in achieving reduction in water use.
The effect of pricing on water use is highly dependent on
local conditions, including the pattern of water use. Studies
show, for example, that a 10-percent increase in price may
affect a reduction in overall use as great as 12 percent or
as little as zero change. In Chicago, a recent study showed
that price changes had no significant impact on use in the
central city but did significantly affect suburban use.-'-'*
The reductions that are achieved may not be long term, for
as real incomes rise, consumption may start to rise again.
Nevertheless, the need for expanding supply is postponed for
a while.
Response to price changes varies with the type of water use.
A 1964 survey of urban areas over 25,000 population indicated
a weighted average water use pattern of 42 percent residential
use, 21 percent commercial, 20 percent industrial, and 17 per-
cent public use (public institutions, street cleaning, etc.).^
Commercial establishments, including laundries and car washes,
are ordinarily responsive to prices charged. Pricing response
for manufacturing water use is highly variable, depending on
the industry and plant design. Since the major component of
industrial use is for cooling purposes (a low-value use), higher
prices would probably have a significant impact. Residential
water use is extremely important in most metropolitan areas,
because it is not only a major use but is generally the greatest
contributor to peak demands on the supply system. Table I
indicates that residential use, primarily lawn sprinkling,
may be many times greater on the peak summer day than the
average through the year.
Residential water use may be divided into in-house or domestic
use and lawn sprinkling. The most comprehensive study of
price effects on residential use was carried out by a Johns
Hopkins University study group in the early 196O's.6 Price
effects, as Table II indicates, were found to vary by type
of use and by region of.the country.
These data indicate that though the effect of price on use
within a household may lie expected to be modest, there is a
significant effect on sprinkling uses, particularly in the
eastern part of the country. Hanke found that the metering
effect in Boulder resulted in (1) significant reductions in
the amount of water used, (2) increased attention to water
leakage, and (3) even a reduction in the area of yard sprinkled.
8
-------
Table I. SUMMARY OF RESIDENTIAL WATER USE*
Type of
study area
Metered public water
and public sewers :
Western states (10 areas)
Eastern states (13 areas)
Metered public water and
septic tanks (5 areas)
Flat-rate public water and
public sewers (8 areas)
Apartments (5 areas)
Total (41 areas)
Average
annual
use"*"
458
310
245
692
191
398
Average
maximum daily
979
786
726
2,354
368
1,096
*Source: Linaweaver, F. P., Jr., et al. "A Study of Resi-
dential Water Use," 1967, HUD TS-12, U. S. Govern-
ment Printing Office, Washington, D. C.
tGallons per day per dwelling unit. /
-------
Table II. EFFECT OF PRICE ON RESIDENTIAL
WATER DEMANDS, 1963-65
Type of use Percent reduction in use after
and area a 10-percent price increase
Domestic 2
Sprinkling 11
Western states 7
Eastern states 16
10
-------
As noted in Table I, water use in apartments is lower than
in detached homes. Similarly, seasonal variations that are
due to sprinkling demands are less for apartments. Therefore,
increases in water prices will have much less impact on the
water use of established apartment areas. However, there is
some evidence that increasing water rates results in greater
use of modified water-using equipment (toilets, showers, and
washing machines that use less water than the unmodified ver-
sions) . The installation of meters and significant increases
in water prices can be expected to lead to some water-saving
practices through the installation of water-saving equipment
and improved maintenance programs.
The practice of using "declining block" water pricing is wide-
spread in metropolitan areas: The more water a user consumes,
the less he pays per unit. This practice of promotional pricing
can encourage inefficient water use. For example, in some
areas, declining block pricing policies actually permit sub-
urban users to pay less per unit of water for lawn sprinkling
at times when they are burdening the supply system with the
most costly peak demands. In such a situation, prices are
lowest when marginal costs are highest. The installation of
meters and the use of cost-based pricing policies will lead
to (1) more efficient use of presently developed water sup-
plies, and (2) the deferral of increasingly costly investments
for development of new supplies.
Verifying the Conclusions Through Data Analysis
To verify some of the conclusions reached by Hanke and others
with regard to the effect of pricing on the consumption of
water, an analysis has been performed on data collected from
EPA's Community Water Supply Survey (CWSS).9 This survey was
initiated in February 1969 by the Bureau of Water Hygiene of
the U. S. Public Health Service (now the Water Supply Research
Laboratory and the Water Programs Division of EPA) in cooper-
ation with state and local health departments and the water
utilities. This study had two purposes: (1) to determine
whether or not the quality of the American consumers' drinking
water met the 1962 U. S. Public Health Service Drinking Water
Standards, and (2) to determine the nature and reliability of
these water supply systems.
The study was designed to assess drinking water quality in
urban areas in each of the regions of the Department of Health,
Education, and Welfare (DHEW). Study areas were selected to
give examples of the several types of water supplies in the
country. The whole Standard Metropolitan Statistical Area (SMSA)
was the basis of the study, with assessments made of all public
water supplies in each study area. This coverage allowed an
assessment of the drinking water quality of the large central
11
-------
city, the suburbs, and the smaller communities located in
the counties in the SMSA and of the interaction between them.
The SMSA's studied in each DHEW Region were: Region I, State
of Vermont; Region II, New York, N. Y.; Region III, Charleston,
W. Va.; Region IV, Charleston, S. C.; Region V, Cincinnati, 0.,
Kentucky, Indiana; Region VI, Kansas City, Mo., Kansas; Region VII,
New Orleans, La.; Region VIII, Pueblo, Colo.; Region IX, San
Bernadino, Riverside, and"Ontario, Calif.
Each water supply was investigated using two criteria: (1) the
water quality delivered to the consumer, and (2) the nature and
adequacy of the water supply system itself. The drinking water
quality was determined by sampling the finished and distribv'-ed
water and returning these samples to the laboratories of the
Bureau of Water Hygiene for physical, bacteriological, and
chemical analyses. The nature and adequacy of the water
supply system was determined by evaluating five items, where
applicable: (1) source(s), (2) treatment, (3) distribution
system, (4) operation and operators, and (5) surveillance pro- • .
gram. This evaluation was accomplished by a field inspection
of the supply system and the gathering of data on three standard
forms. In addition, water quality data for the previous year
was obtained from state and county health department records.
When these investigations were completed, each water supply
was rated either as excellent, good, fair, poor, or unfit to
drink.
In addition to information related to the quality of service,
data was collected on the cost of operation arid quantities of
water consumed. It is this data that has been analyzed using
the Cincinnati SMSA as an example.
The Cincinnati SMSA-—
The Cincinnati SMSA includes Dearborn County in Indiana, Boone,
Campbell, and Kenton Counties in Kentucky, and Clermont,
Hamilton, and Warren Counties in Ohio. Data were collected
as a part of this survey on 67 public water supplies, including
57 community water supplies (those that piped wat^r to homes)
and 10 special water supplies (those that served institutions,
trailer parks, and water haulers).9 Twenty-two of these com-
munity water supplies were analyzed to relate factors affecting
the cost of water delivered to the consumer to physical factors,
and the price of the consumer's water to his utilization.
Table III summarizes the statistics for the water supplies
selected for analysis.
Two approaches were attempted in the analysis of the data.
The first analysis was conducted in an attempt to determine
the factors that relate the cost of water supply systems to
the cost of supplying the finished product. In this case,
12
-------
Table III. SUMMARY OF DATA FROM WATER SUPPLIES SELECTED FOR ANALYSIS
OJ
-
City
Cincinnati
Franklin
Indian Hill
Norwood
Reading
Campbell
Covington
Florence
Newport
Harrison
Lebanon
Love land
Mason
Wyoming
Aurora
Lawrenceburg
Ludlow
Glendale
Milford
South Lebanon
Williamsburg
Highland Heights
Pop.
served
(1)
850,000
11,000
4,526
31,000
15,000
44,000
64,000
15,000
30,070
5,140
6,500
5,000
6,200
10,000
6,300
5,200
6,010
3,000
4,530
2,720
2,200
4,000
Central
city pop.
density
(persons/
m2)
(2)
6,569.3
4,166.8
255.0
11,526.7
4,752.6
3,942.9
10,062.7
2,012.8
10,739.3
4,308.9
4,308.9
2,276.4
1,524.8
2,865.2
168.7
2,382.9
8,904.3
1,660.6
4,131.0
4,533.3
1,656.2
2,909.2
Average
daily
demand
(106 gal)
(3)
112.60
1.50
1.80
4.00
1.85
2.43
5.46
0.75
3.39
0.36
1.05
0.57
0.25
0.90
0.517
0.25
0.335
0.30
0.44
0.32
0.10
0.13
Water
sup-
plier*
(4)
1
1
1
0
1
0
1
0
1
1
1
1
1
1
1
1
0
1
1
1
1
0
Quality
indi-^
Water
source'''
(5)
1
0
0
1
0
1
1
1
1
0
0
0
0
0
Q
0
1
0
0
0
1
1
cators-f
Q!
1
1
1
1
1
1
1
1
1
1
0
1
1
1
1
1
1
1
1
0
1
1
Q2
(6)
1
0
1
1
0
1
0
1
1
0
0
0
1
1
0
0
0
0
0
0
0
1
Cost/
base
price
($/l,000
gal)
(7)
0.35
0.45
0.65
0.30
0.40
0.70
0.90
1.30
0.75
0.70
0.30
0.50
1.15
0.80
0.85
0.70
0.40
0.70
0.55
0.20
1.15
0.70
Per
• • capita
income
(8)
$ 3,141
2,813
10,268
3,148
3,171
2,796
2,476
3,441
2,308
2,902
3,098
2,925
3,159
5,974
2,710
2,639
2,590
4,778
3,296
2,544
3,543
4,225
Average
water
use
(gal/
capita/
day)
(9)
132
136
120
129
123
55
85
50
97
71
162
76
40
90
82
48
55
100
97
118
45
32
*1 = Primary source; 0
tl = Surface supply; 0 = ground supply.
+1,1 = Good; 1,0 = fair; 0,0 = poor.
-------
the data in Columns 2 through 8 were regressed against the
data in Column 9. The second analysis is an attempt to re-
late the price of water to consumption, and the data in
Columns 9 and 10 were regressed against the data in Column 11.
The cost analysis will be discussed initially.
Cost Analysis—The variables in this analysis are defined as
follows:
• C = the base cost in dollars per 1,000 gal. of water;
« Ps = the population served by the water utility;
• PD = the population density in persons per square mile for
the service area;
• D = average daily demand for water in millions of gallons;
• Sp = supplier of water, where Sp = 1 implies that the
utility is the primary supplier of water/ and Sp = 0 implies
that the utility is a secondary supplier;
• So = source of the water the utility supplies, where So = 1
implies that the utility uses surface water, and So = 0 implies
that the utility uses ground water; and,
• QI and Q2 are dummy variables indicating the rating of the
water utility, where Qi = 1, Q2 = 1 implies a rating of good,
Ql = 1, Q2 = 0 implies that the rating is fair, and Q]_ = 0,
Q2 = 0 implies that the rating is poor.
The analysis is static or cross-sectional; time variations are
not considered, but several points in the same period are
studied. By this approach, changes over time can be ignored,
and the differences between places can be emphasized. In
addition, by choosing a single metropolitan area, differences
in the price level of equipment, labor, and capital are emphasized,
Using a step-wise regression analysis, we obtain the following
equation:
C = 0.00000944 Ps - 0.00005972 PD - 0.07722744 D
(1.582008) (-4.2185103) (-1.73036672)
+ 0.42094866 Sp + 0.47268092 So + 0.26372458 Qx
(3.1656385) (3.352057) (1.95385515)
+ 0.15042853 Q2 + 0.1024940 .... (1)
(1.66660721)
14
-------
The multiple correlation coefficient for this equation is
approximately 0.81, which means that nearly 66 percent of
the variation of the data is explained by this equation.
The t values for each variable are shown in parentheses below
that variable. Based on this analysis, it is apparent that
such variables as population density/ whether or not the
utility is a secondary or primary supplier, and the source
of water play the primary roles in determining the cost of
a water supply system. Total population served, total demand,
and quality of the system play secondary roles.
In Equation 1, as population density decreases (negative value),
cost per 1,000 gal. increases proportionally. This implies
significant economies of scale with density and is consistent
with the findings of other investigators. As might be ex-
pected, the cost of the supply is influenced by the source
of water. Surface supplies are the most expensive, no doubt
because of the storage required when surface water is used.
Most interesting is the significance of the supply variable Sp,
which indicates that secondary suppliers charge less for water
than the utilities from which they buy it. This practice
appears to be contrary to common sense until one examines the
management practices among water utilities. Large users of
water commonly receive discounts, and such is apparently the
case in this analysis. The small utilities that buy water
from the large utilities pay less per unit for their water
than the large utilities' individual consumers. The small
utilities then pass this discount along to their own customers.
Price Versus Consumption—In this analysis, data from Table III,
Column 9, were regressed against data from Column 11, with
price as the independent variable and consumption as the
dependent variable. Several different models were used in
an attempt to relate price (cost per gallon) versus consump-
tion (gallons per capita per day). The models examined were
as follows:
• Linear: Y versus X;
• Log-Log: LnY versus LnX;
• Inverse: Y versus 1/X;
• Semi-Log: LnY versus X; and,
• Inverse Semi-Log: LnY versus 1/X.
In the following equations, P = price, and C = consumption.
The t statistic is given in parentheses below the independent
variable, and R2 is given for each equation. The equations
are as follows:
15
-------
• Linear: C = - 85.32896P + 144.5577, R2 = 44.512 ... (2)
(-4.2244442)
• Log-Log: LnC = - 0.6019963LnP + 1.7718307,
(-3.7609605) R2 = 38.497 ... (3)
• Inverse: C = 22.605648/P + 45.714695., R2 = 37.751 . . (4)
(3.7061572)
• Semi-Log: LnC = - 0.4348467P + 2.1947295,
(-3.8130064) R2 = 39.199 ... (5)
• Inverse Semi-Log: LnC = 0.1129775/P + 1.6934253,
(3.257656) R2 = 31.400 ... (6)
There is little to indicate that one model yields superior
results when compared to another. Price elasticity of demand
when calculated from the Log-Log model is -0.602. When calcu-
lated by expanding around the mean of the linear model, price
elasticity is:
dQ . P
Q
e = -0.63
Both of these results are consistent with each other and with
intuition*. These elasticities indicate that water consumption
is inelastic (e < 1) but that they are significantly different
from zero. By increasing the price by 10 percent, consumption
would be expected to be reduced by 6 to 6.3 percent. These
results are consistent with those reported by Hanke; but more
importantly, they are based on data that have been accumulated
uirectly from municipal water supplies in a cross-sectional
study. The consumption patterns in this metropolitan area
should have reached steady state as opposed to Hanke's work
and should, therefore, reflect the true reaction of the con-
sumer to price.
To extend this analysis, income per capita (Column 10 in
Table III) is added as an independent variable to Equation 2
in the analysis above. If I = per capita income, the fol-
lowing relationship results:
C = - 87.688024P + 0,00429871 + 130.882474, R2 = 46.097 . . (7)
(-4.385257) (1.2601998)
16
-------
From this analysis, it appears that per capita income has
little effect on the consumption of water. This may be true
for a cross-sectional analysis, but over time and over a
broader geographical area, per capita income may be signifi-
cant.
PRICE AND ITS IMPACT ON TECHNOLOGY
The impact of the price of water on water supply design is
extremely important. Linaweaver^ developed a design equation
based on expected average domestic water use, irrigable land,
potential evapotranspiration, and potential precipitation.
This equation is as follows:
Q = Qd + 0.6 c a Ls(Epot - Peff)
.#*••
where:
Q = expected average demand for any period expressed as a
rate in gallons per day;
0.3 = expected average domestic (household) use in gallons per
day, which applies for all periods of a day or longer and may
be reliably estimated from a simple function of the average
market value of dwellings as described below;
0.6 = a coefficient to adjust for the difference between
actual evapotranspiration from lawns and potential evapo-
transpiration ;
c = constant to adjust for units, 2.72 x 104 gal. per acre-
inch of water;
a = number of dwelling units;
e*
Ls = average irrigable area in acres per dwelling unit;
%>ot = estimated average potential evapotranspiration for the
period of demand in question in inches of water per day; and,
•Peff = amount of natural precipitation effective in satisfying
evapotranspiration for the period and thereby reducing the
requirements for lawn sprinkling in inches of water per day.
The principal quantity considered for design is the expected
maximum daily demand, which is likely to occur during the
latter part of June or during July, when high temperatures
and long hours of sunlight combine to make a high evapotran-
spiration. Typical values can be assumed for the constants
in Linaweaver's equation. These are listed as follows:
17
-------
c = 2.72 x 10
= °'20
L = 0.28
s
Qd = 244
Q = 1/160 gpd per dwelling unit
max 3jr c
If a 10 percent increase in price is assumed based on the
previous analysis, a 6 percent decrease in consumption would
be expected. However, this price elasticity is based on the
mean value for consumption, and it is likely that the reduction
in consumption for lawn sprinkling and other uses have a greater
price elasticity than 6 percent. Hanke shows a price elasticity
of 1.12 for residential sprinkling. If the price elasticities
were separated and the 6 percent figure were applied to Qd and
the 11.2 percent to the sprinkling demand in Linaweaver's
equation, a reduction would occur in Q to a value of 1,043 gpd.
If this change in the demand for system capacity were actually
to take place with an increase in price, it would be expected
to cost the consumer more money, since the data show that the
situation is relatively inelastic. However, it also has some
long-term benefits, since it would also increase the life of
existing water resources and save in long-term capital invest-
ment. Additional research might be fruitful on these types of
trade-offs.
PRICING AND ITS IMPACT ON WATER REUSE
The problem of supplying water for municipal use is becoming
increasingly difficult in the United States, and this trend
is expected to continue in the foreseeable future. According
to recent estimates of the Water Resources Council, several
regions in the United States may experience critical problems
of water shortage and water quality deterioration unless
suitable measures are taken. Under certain conditions, the
optimal solution to increasing municipal demands and reducing
water quality deterioration may rest with wastewater renova-
tion. Advanced waste treatment could lower the amounts of
pollutants discharged to prescribed levels, and the resultant
high quality effluent could be made available on site for
municipal use.'
The use of renovated wastewater provides an additional source
of water supply, but it provides what would be termed an
engineering solution to the problem. Municipal demands could
also be met more effectively by changing the way in which
18
-------
water is used. This can be accomplished by reducing the
amount of water used or altering demands, as has been dis-
cussed earlier.
Reduction in use could be accomplished by metering or pricing
policies, use restrictions, or recycling. Both metering and
pricing policies have been found to be an effective means of
lowering municipal water demands, although they are not (at
this time) employed by water managers for this purpose. Sev-
eral major cities, including New York and Chicago, do not
meter their water supplies but instead, charge a flat rate.
The practicability of metering must be evaluated according
to the costs of installation and operation versus the in-
creased supply that would be available as the result of their
installation. One study discussed earlier indicated a 40 per-
cent drop in per capita water consumption in Boulder, Colo.,
from 1960, when only 5 percent of the city was metered, to
1965, when the city was fully metered. Another study indi-
cated that the quantity demanded for residential uses, par-
ticularly sprinkling, is affected by the price charged. It
has been suggested that even maximum-day demands respond to
price changes, a fact that could be used by management for
altering average and maximum-day demands.
The new Federal Water Pollution Control Act requires that no
pollution be discharged to the Nation's streams by 1980. The
achievement of this goal will be extremely expensive, and the
quality of effluent that might result will be extremely high.
It makes sense, therefore, from an engineering point of view,
that these effluents be recycled for municipal use. If eco-
nomic measures (pricing) were applied to the original drinking
water stream reaching the consumer, the amount of domestic
water consumed andt therefore, the amount of water discharged
to be treated, will be reduced. These lower effluent quanti-
ties will have higher concentrations of pollutants, a situation
that will have an impact on the kind of technology used for
treatment. The higher concentrations of waste will probably
make the treatment process more efficient. At any rate, by
introducing a pricing policy for water, there should be sub-
stantial changes in technology. These changes may actually
reduce treatment, collection, and distribution costs.
This approach will tend to make the cost of water supply and
sewage services explicit and, hopefully, more efficient. At
this point, these ideas are only speculation, but they provide
some interesting thought for joint exploration by economists
and engineers.
19
-------
SUMMARY AND CONCLUSIONS
The current state of knowledge with regard to the pricing of
water supplies and its impact on water consumption has been
examined in some detail. Although considerable work has been
done by economists in this area, little of the information
generated has been used or possibly understood by practicing
engineers. In an attempt to demonstrate that the idea of
pricing is not a wild-eyed economic theory/ engineering data
collected by the EPA have been used to support the basic
premise of pricing and its impact on water utilization.
These data were then analyzed for their relation to the cost
of supply and pricing for consumption, and some of the impacts
that pricing might have on water supply technology were sketched.
It was also suggested that the pricing of water supplies to
reduce consumption could have a significant effect on the
conservation of potable supplies.
Several obvious conclusions arise from this analysis:
• Pricing should be considered in conjunction with technology
for the design of water supply systems.
• There are significant diseconomies of population density
in water supply systems. That is, as densities decrease, the
average per capita cost for water increases.
• Average price elasticities of demand for water are less
than one but significantly greater than zero.
• The pricing mechanism could have a significant impact on
water supply technology.
• Pricing should be considered in conjunction with technology
when considering the issue of water reuse.
• Management practices in which a large utility gives dis-
counts to smaller utilities leads to price discrimination
against its own customers.
20
-------
REFERENCES
1. Craine, Lyle C. , "Water Management Innovations in England,"
Published for Resources for the Future, Inc., Johns Hopkins
Press, Baltimore, Maryland (1969).
2. Hanke, Steve H. , "Demand for Water Under Dynamic Conditions,"
Water Resources Research (5), October 1970.
3. Hanke, Steve H., "The Theory of User Fees and Its Applica-
tion to Water," Public Prices for Public Products, S. J.
Mushkin, ed. , The Urban Institute, Washington, D. C. (1971).
4. Herschel, Clemens, "Frontinus and the Water Supply of the
City of Rome," Longmens, Green and Company, New York (1913).
5. Hittman Associates, Inc., "Price, Demand, Cost and Revenue
in Urban Water Utilities," HIV-474, Hittman Associates,
Inc., Columbia, Maryland (1972).
6. Howe, C. W., and Linaweaver, F. P., Jr., "The Impact of
Price on Residential Water Demand and Its Relation to
Systems Design and Price Structure," Water Resources
Research 3(l):13-32 (1967).
7. Johnson, James F., "Renovated Waste Water," The University
of Chicago, Department of Geography, Research Paper No. 135.
8. Linaweaver, F. P., Jr.; Geyer, John C., and Wolff, Jerome
B., "Final and Summary Report on the Residential Water Use
Research Project," The Johns Hopkins University, Environ-
mental Engineering Science, Baltimore, Maryland, June 1966.
9. Maddox, Franklin D., "Water Supply and Consumer Costs,
Cincinnati Standard Metropolitan Statistical Area," an
unpublished report for the Bureau of Water Hygiene, March 5,
1970.
10. Murray, C. Richard, "Estimated Use of Water in the United
States," 1965, U. S. Geological Survey Circular 556, U. S.
Department of the Interior, Washington, D. C. (1968).
11. National Water Commission Act, P. L. 90-515, September 26,
1968, 82 Stat. 868, 42 U. S. C. A. (1962).
21
-------
12. National Water Commission, Proposed Report of the National
Water Commission, Review Draft, Washington, D. C., November
1972.
13. Whitford, Peter W., "Forecasting Demand for Urban Water
Supply," Report EE, Stanford University, Palo Alto, Cali-
fornia, p. 36 (1970).
14. Wong, S. T., "A Model on Municipal Water Demand: A Case
Study of Northwestern Illinois," Land Economics 48(1),
pp. 34-44, February 1972.
22
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
. REPORT NO.
EPA-670/1-74-001
2.
3. RECIPIENT'S ACCESSIOWNO.
4. TITLE AND SUBTITLE
Pricing for Water Supply; Its Impact on
Systems Management
5. REPORT DATE
April 1974—Issuing Date
6. PERFORMING ORGANIZATION CODE
. AUTHOR(S)
Robert M. Clark and Haynes C. Goddard
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
National Environmental Research Center
U. S. Environmental Protection Agency
Cincinnati, Ohio 45268
10. PROGRAM ELEMENT NO. lCAU4b
ROAP-51ASB, Task 01
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Same as above.
13. TYPE OF REPORT AND PERIOD COVERED
In-House
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT "~~~
Problems related to water supply have become increasingly important
in recent years. In the past, water has been so abundant that it was
available in almost unlimited quantities. But this is no longer the
case in many parts of the United States. Water has become a resource
that is relatively scarce. And the land, labor, and capital resources
needed to convey water to places of useful application and to collect
and treat wastewater are also scarce.
This paper discusses current pricing policies by water utilities
and the changes in consumption patterns which other investigators
have found resulting from the changes in price for water supplies.
Water consumption and pricing data are analyzed for a specific SMSA.
The potential impact of pricing policies on technology and wastewater
reuse is discussed.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
COS AT I Field/Group
economic analysis, economics,
water, water consumption, water
quality, water supply, cost
analysis, regression analysis
pricing policies,
Standard Metropol-
itan Statistical
Area, wastewater
reuse, peak-load
problems, resource
allocation
05/C
13/B
8. DISTRIBUTION STATEMENT
Release to public,
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
27
23
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
*U.S.Government Printing Office: 1974 — 757-581/5313
------- |