WATER POLLUTION CONTROL RESEARCH SERIES • 16110 DTP 02/71
The Economics
of
Water Supply and Quality
ENVIRONMENTAL PROTECTION AGENCY • WATER QUALITY OFFICE
-------
MATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Series describes
the results and progress in the control and abatement
of pollution in our Nation's waters. They provide a
central source of information on the research, develop-
ment, and demonstration activities in the Water Quality
Office, Environmental Protection Agency, through inhouse
research and grants and contracts with Federal, State,
and local agencies, research institutions, and industrial
organizations.
Inquiries pertaining to Water Pollution Control Research
Reports should be directed to the Head, Project Reports
System, Office of Research and Development, Water Quality
Office, Environmental Protection Agency, Room 1108,
Washington, D. C. 20242.
-------
The Economics of Water Supply and Quality
by the
Harvard Water Program
Harvard University
Cambridge, Massachusetts
for the
WATER QUALITY OFFICE
ENVIRONMENTAL PROTECTION AGENCY
Project 16110 DTF
(Formerly WP-00919)
February 1971
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, B.C. 20402 - Price 50 cents
Stock Number 5501-0070
-------
EPA Review Notice
This report has been reviewed by the Water
Quality Office, EPA, and approved for publication.
Approval does not signify that the contents
necessarily reflect the views and policies of
the Environmental Protection Agency, nor does
mention of trade names or commercial products
constitute endorsement or recommendation for
use.
-------
ABSTRACT
Research has been directed to a series of problems in water resource
management. One pair of projects has focused on problems of water
supply planning under uncertainty, and improved methods of analysis
for planning have been developed.
A set of four research efforts has been devoted to problems of water
quality management, with special attention given to the decisions
faced by municipalities and regional authorities. New conceptual
models of quality management institutions have been proposed, and
the conceptual framework has been applied to several sample river
basins.
A third focus has been on the development of improved computing methods
for analysis of river basins. Two projects have contributed new analyt-
ical models for preliminary screening of river basin designs. Special
attention has been directed to the analysis of flow augmentation for
water quality management.
Finally, research has been carried out on the use of regional analysis
in water resource planning. A static regional input-output model has
also been developed that can analyze the interrelations between regional
growth and efforts at water quality control.
As a subsidiary result of this grant research, a monograph on water
quality management has been prepared. It includes much of the material
described above, as well as five additional essays.
This report is submitted in fulfillment of Grant Nos. WP-00919 and 16110
DTP from the Water Quality Office, Environmental Protection Agency.
-------
TABLE OF CONTENTS
Section Page
I Introduction 1
II Research Efforts Undertaken 3
Research on the Economics of Water Supply 3
—Municipal Water System Planning 3
—Theoretical Models of Response
to Uncertainty 6
Research on the Economics of Walter Quality 9
—Influence of Technical, Economic, and
Political Factors on Quality Manage-
ment Efforts 10
—Municipal Waste Management 13
—Municipal Evaluation of Alternative
Quality Management Plans 15
—Estuarine Water Quality Management 19
Research on Computing Techniques for
Basin-wide Planning 20
—Interaction of Screening Models and
Basin Simulation 20
—Systems Models for Evaluation of
Flow Augmentation 22
Research on Use of Regional Analysis in
Water Resource Planning 25
—Regional Analysis of Quality Control
Costs 25
—A Model of the Interdependence between
a Regional Economy and Water Quality
Control Efforts 26
III A Monograph on Water Quality Management 29
IV Acknowledgments 33
V References 35
-------
SECTION I
INTRODUCTION
This report covers four years' research by members of the
Harvard Water Program.1 The greater part of the financial
assistance for this work on "The Economics of Water Supply
and Quality," was provided originally by the U. S. Public
Health Service2 (Grant No. WP-00919). The grant was for a
three-year period, January 1966 through December 1968. The
terms and level of financing remained unchanged when, in
1966, the grant program in water quality was transferred
from the U. S. Public Health Service to the Federal Water
Quality Administration of the U. S. Department of the Interior.
This report summarizes work performed during the initial period,
and includes research supported by a nine-month extension that
began in January 1969 (Grant No. 16090 DTP).
Section II of this report presents a summary of the work ~
carried out by the different participants in the program.
Although the research described was supervised by Robert
Dorfman, Harold A. Thomas, Jr-, and Henry D. Jacoby, a
major portion of the work was carried out by undergraduate
students, graduate students, and visiting scholars.
p
Financial support was also contributed by Resources for the
Future, Inc., Washington, D. C.
D
JThe full reports of research results are contained in a s.et
of Harvard Water Program Discussion Pape-rs, various working
documents, and a number of published articles and books.
-------
It is organized according to the four principal areas of
research outlined in the original grant proposal:
—the economics of water supply,
—the economics of water quality,
—computing techniques for basin-wide planning, and
—the use of regional analysis in water resources
planning.
In addition to research in these four areas, a subsidiary
purpose of Grant No. 16090 DTP was to facilitate the prepara-
tion of a monograph on water quality management. The mono-
graph is titled Models for Managing Regional Water Quality.
Six chapters of the completed monograph are the direct
result of research financed by Grant Nos. WP-00919 and
16090 DTP. Another two chapters are the result of Investi-
gation done under grants from the FWQA to re'searchers at
Harvard University for work on "Operations Research in Water
Quality Management" (Grant No. WP-00111) and "Dynamic Planning
Techniques for Pollution Abatement" (Grant No. 16090 DSS).
-------
SECTION II
RESEARCH EFFORTS UNDERTAKEN
Research on the Economics of Water Supply
Many of the water-related problems of public policy Involve
systems subject to uncertainty In supply—e.g., municipal
water systems subject to uncertainty in climate and in con-
sumer demand, or water quality systems which are influenced
not only by fluctuation in river flow and temperature but by
variation in pollution loads. All planning models must deal
with this uncertainty in one fashion or another, and in our
work we have been particularly concerned with the economic
benefit and loss functions that are used to represent the
economic impact of variation in water supply and quality.
Two of our research efforts were devoted to aspects of planning
under uncertainty. Both were oriented primarily toward
problems of water supply: one (introduced first below) drew
on the experience of water systems in deficit in order to
study the economic losses associated with water shortage and
to analyze potential improvements in water supply planning
practice; the other was primarily theoretical in nature and
explored the expected responses of different types of economic
units to uncertainty in supply.
Municipal Water System Planning
The Problem of Drought and Water Supply Shortage
During the early months of the grant period, our interest in
the economics of water supply and the problems of system
planning was heightened by the fact that the northeastern
United States was passing through what later turned out to be
the most serious period of a prolonged and severe drought.
Many municipal water systems experienced shortage during
this period, and the event provided a rare opportunity for
-------
research on the economic loss associated with water supply
restrictions. That so many systems experienced difficulty
also raised a question about the nature and adequacy of tht
planning procedures used in the design of municipal water
supplies. Research on this problem was carried out by
Clifford S. Russell, and the results are reported in The Recent
Northeast Drought: Short-run Costs of Water Shortage and
Lessons for Water System Planning [15J . ^
The Russell study contributes insight into the link between
system adequacy, climatic variation and the levels of
shortage experienced. And this insight, in turn, leads to
a method for making probability statements about various
levels of shortage for given system adequacy levels. The
study also produces evidence on the economic impact of water
shortage, which makes it possible to translate shortage
figures into costs that can be compared with the costs of
system improvements. These results lead to the formulation
of a revised method for water system planning.
The framework within which the model is presented is that of
a capacity expansion model in which the costs of meeting a
growing demand are balanced against the losses incurred by
not doing so. Attention is concentrated on additions to sys-
tem size (as measured by the amount of water available at a
given level of assurance) as the relevant method of improving
system adequacy. The alternative possibility of altering
system demand or of making more efficient use of the water
available from a given system (as, for example, with leak-
control programs) is outside the scope of the study. In
addition, it is assumed that the historically given price
structure is to be continued indefinitely into the future.
All alternative plans for the growth of the system are
assumed to share the aim of meeting the level of demand
implied by that given price structure.
In order to discuss the model and the drought study findings,
it is necessary to introduce a few key definitions.
(i) System Capacity (or Safe Yield) is that constant
rate of withdrawal from a surface water system which is
assured at some chosen probability level (e.g., that constant
withdrawal rate which can be maintained 95 percent of the time
or in a repetition of the "five percent drought").
In this effort, we were fortunate to have the collaboration
of Professor Robert Kates and Dr. David Arey of Clark
University who also were involved in a study of the impact
of the drought under the sponsorship of Resources for the
Future, Inc. The results of our two research efforts have
now been combined into a book by Russell, Arey, and Kates,
Drought and Water Supply: Implications of the Massachusetts
Experience for Municipal Planning_[l6].
-------
(ii) System Demand is the level of demand for withdrawals
from the system which exists under the particular conditions
(e.g., price, weather) being discussed. Total demand may
usefully be thought of as the product of the level of per
capita demand and the size of the population, and demand may
be projected into the future for planning purposes using
separate projections of these variables.
(iii) System Adequacy is indicated by an index
developed in the course of the study: it is equal to the
ratio of (projected) demand to safe yield. The larger this
index number, the lower the level of relative system adequacy.
(iv) Shortage is defined as the percentage difference
between the existing level of demand and the 'actual deliveries
(or amount of water available for delivery, in the future).
(v) Climatic Variation is relevant both to the level of
demand and of supply. Most attention, however, is focused
on the effect of climate on the availability of water. The
principal source of information on climatic variation is
rainfall data.
Development of a Model for Water System Planning
The construction of a model for the planning of system capacity
expansion involves three major questions. First it is necessary
to know in quantitative terms how the level of system adequacy
and the severity of climatic events interact to create water
shortages. This question is at the heart of the attempt to
attach probabilities to shortage levels. A second question
concerns the size of the economic losses associated with
various levels of shortage. An answer to this question is
necessary in order to translate the analysis of the physical
world into economically meaningful terms. The third question,
central to the model building effort, is that of the costs of
improving system adequacy. Because the study concentrates
on additions to system capacity as the means of improving
adequacy, the third question is restated to involve the cost
of adding to capacity—i.e., of adding safe yield.
As for the first question, the study begins by hypothesizing
a model in which the level of shortage (for a given adequacy
level) depends on the fraction of safe yield flow available,
and where this in turn is determined by the extent to which
current natural inflow of water differs from the flow asso-
ciated with safe yield. The climatic variable used in
measuring the severity of events is the four-year cumulation
of annual rainfall amounts. This series is chosen in prefer-
ence to one of streamflow or watershed yield primarily on
practical grounds of data availability. Investigation, how-
ever, confirms that the series is fairly closely related to
streamflows.
-------
The second question mentioned above, that of the size of the
economic losses caused by varying levels of shortage, is
examined using extensive data collected in three Massachusetts
towns at the end of the 1961-65 drought. These data consist
primarily of interview results from contacts made with local
firms in the commercial and industrial sectors. These are
supplemented by interviews with certain municipal officials,
by newspaper searches and by the use of published results
concerning household demand functions for water.
Regarding the third question, two independent approaches are
used in estimating the costs of adding to safe yield. First,
on the basis of published data a "theoretical" cost function
is estimated. This involves estimating the cost of building
a reservoir of specified volume and then calculating the
storage volume needed on streams of various sizes to provide
a given firm withdrawal rate with (approximately) a five-
percent failure rate. The second estimation method involves
the use of historical cost estimates from engineering reports
submitted over the years to five Massachusetts cities by con-
sulting engineers.
These three aspects are then combined into a planning model
for municipal supply system for which demand is growing over
time. At any particular time demand bears some particular
relation to safe yield, and this implies some expected loss
(considered as a best estimate of the loss to be incurred
during the period). System adequacy may be changed one or
more times during a planning period by making additions to
safe yield. The task of the planning model is to determine
the optimal sizes of the increments to system capacity and
the optimal times to build them over a sixty-year planning
horizon. The resulting model is a nonlinear programming
problem with linear constraints, and its solution is approached
by the "method of feasible directions."
Prom the solutions to this model, it proves possible to
extract information on the optimal size and timing of capacity
additions for any municipal system. And thus this particular
aspect of our research has led to results that may be of
immediate and practical usefulness to water system planners
—both in the design of independent municipal supplies and
in the development of economic data for systems analysis of
larger, interdependent, basin-wide planning problems.
Theoretical Models of Response to Uncertainty
Because the response of water users to variation in supply
—and instream quality—is such an important aspect of the
economics of water system planning, it was decided to seek
an improved theoretical basis for our work in this area.
-------
To this end we took on a study of the behavior of consumers,
business firms, and public decision bodies under uncertainty.
The results of this work, most of which was carried out by
Stephen J. Turnovsky, are reported in a series of working
papers and articles, and in two chapters in the new monograph
on water quality management.
A Model of the Response of Consumers to Uncertainty
in Supply
The main portion of the study focuses on the response of the
individual consumer to situations where the price of a par-
ticular good may be fixed but the quantity actually available
to the consumer is uncertain; furthermore -, attention is con-
centrated on the case where the consumer lacks facilities
for storing the good in question. Uncertainty in water
supply from a stochastic stream or from a badly managed
municipal water system are obvious examples of this type of
situation. The administrative procedures for setting water
rates usually do not allow prices to respond to variation in
supply, and some other rationing device is necessary to allo-
cate the available supply among users. As a result, the
uncertainty in climate or in system performance is trans-
formed into an uncertain supply to the customer.
The initial efforts are directed to the individual consumer,
and the analysis is based on the assumption that the producers
will attempt to provide whatever the consumers plan to purchase,
although the actual supply can only be guaranteed within some
random error of the planned consumption. The theoretical
model of the consumer's response in this situation involves
the extension of the classical utility framework to encompass
stochastic supplies, and an attempt is made to predict how
consumption plans would respond to changes in the variance
of the supply distributions using this model.
The results of this work are reported in a paper entitled "A
Model of Consumer Behavior Under Conditions of Uncertainty in
Supply" [19]. One of the more interesting conclusions is
that, if x. is the planned consumption of commodity i and
0? is the supply variance of commodity i, then
K
8x.
As uncertainty (as measured by supply variance) increases,
consumers plan to purchase more of the commodity. This
result follows essentially because of diminishing marginal
utility considerations which imply t-hat the loss associated
with being short of a particular commodity exceeds the gain
-------
associated with an over-supply. In other words, under uncer-
tainty the consumer increases his margin of safety and orders
more than he would if he were certain of receiving the
desired amount. In addition, the model indicates that a con-
sumer who has already experienced uncertainty in the past
will react less positively to increases in 02i than a con-
sumer experiencing uncertainty for the first time.
An Empirical Test of the Theory
In order to test these theoretical results, the model is
applied to data from the water supply systems of nineteen
Massachusetts towns over the 1961-65 drought period. Demand
functions for water are estimated using a combination of
cross-section and time series analysis and drawing on data
made available by the work of Clifford S. Russell discussed
above [15]. Both the phenomena predicted by the theoretical
model are found to exist in this particular case: a| proves
to be a significant explanatory variable in the estimated
demand equations, but less important when the consumer has
been subjected to uncertainty in the recent past. This
research is described in detail in "The Demand for Water:
Some Empirical Evidence on Consumers' Response to a Commodity
Uncertain in Supply" [24].
Extension of the Model to Consider the
Behavior of the Firm Facing Uncertain
Factor Supplies
The same theoretical apparatus which is applied to the consumer
is also turned to the problem of the firm with uncertainty in
factor supplies. The firm studied is assumed to produce a
single output and to trade in competitive input and product
markets. But as opposed to the conventional analysis of the
competitive firm, the quantities of input factors available
are assumed to be uncertain. Once again prices are assumed
fixed. The firm is assumed to decide what inputs it plans to
use and to convey these plans to the suppliers of raw
materials (e.g., water) and other inputs. The suppliers of
inputs then undertake to provide the desired amounts, but due
to hydrologic uncertainty are only able to guarantee their
supplies within some random error, so that the quantities
.actually delivered to the firm are in general different from
those which it anticipates.
The behavior of the competitive firm under these circumstances
is explored in a paper on " The -Theory of Production Under
Conditions of Stochastic Input Supply" [23]. A similar case
where the firm can buy any quantity it desires of the input
factor but where the input price is uncertain, is analyzed in
8
-------
a paper on "The Behavior of a Competitive Firm with Uncertainty
in Factor Markets" [20]. Still another application of this
same model is the case where the firm is faced with uncer-
tainty in demand for its product. In "Stochastic Demand and
the Theory of the Firm" [21] an attempt is made to delve
beneath the firm's revenue and cost functions and to analyze
how decisions concerning the use of inputs in the production
process are affected by the uncertainty in demand for the
final product.
Government Investment Under Uncertainty
In "Some Criteria for the Optimal Selection of Risky
Investment Projects" [22] the models developed in the pre-
ceding papers are applied to the general problem of Invest-
ment under uncertainty: attention is focused on the problem
of optimal government investment in projects whose outputs
—which are services or commodities consumed by the population
—are uncertain. As before, the example.used is water:
because of uncertainty caused by natural phenomena, the
government supplier cannot guarantee 100 percent reliability
and can only undertake to satisfy consumers' demand with some
prescribed probability. The parameters that describe the
probability distribution of supply experienced by the con-
sumers reflect, at least in part, the design of the water
system, and to that extent they are within the government's
control. For the purpose of this analysis, it is assumed
that these parameters are completely within the government's
control, and that by increasing expenditure an improved
systems can be provided. Accordingly, once it has been
established how consumers may be expected to respond to—
and are affected by—uncertain supply, the next step is to
determine the socially optimal design*
Many of these research results were drawn together into
two essays focused specifically on the problem of uncer-
tainty as it affects the design and operating decisions
made by public agencies [25, 26].
Research on the Economics of Water Quality
Our work on the economics of water quality has placed
heavy emphasis on the interaction of economic, technical,
and political factors in water quality management problems.
Particular attention has been devoted to the regional
quality management authority as a key agent in many quality
management efforts, and to the municipal entities which they
must regulate. Our research also has touched on special
problems of estuarine quality management.
-------
Influence of Technical, Economic, and Political
Factors on Quality Management Efforts
One of the central research efforts, conducted by Robert
Dorfman and Henry D. Jacoby, has been directed to the
study of political and economic considerations as they
impinge on technical and engineering solutions of water
quality problems. This work emphasizes that what ought
to be done is never clear, and is a matter of judgment,
about which opinions differ legitimately. What will be
done about water quality in any context is therefore a
reconciliation of diverging opinions and conflicting
interests. This obvious fact would not be worth mentioning
but for two circumstances. First, this conflict is almost
universally ignored (or at least concealed) in most discus-
sions of water resource design and project evaluation.
Second, as a consequence of the first, there are no known
methods of design that take political considerations into
account.
The Task of River Basin Authorities
In the attempt to study the institutional and political side
of water quality management, attention has been focused on
the most promising legal expedient for dealing with quality
problems—the River Basin Authority. A River Basin Authority
is a special governmental institution empowered to regulate
the use of a segment of a river, or in some cases an entire
river. The powers to be conferred on River Basin Authorities
to enable them to discharge their functions are still obscure.
At the very least, a River Basin Authority is usually respon-
sible for prescribing levels of waste treatment for individual
dischargers or setting limits on the quantities of pollutants
that may be discharged. It also may be authorized to levy
various taxes and charges to finance its operations and to
induce water users to behave in ways that make efficient use
of available assimilative capacity. In some cases, an
Authority may be empowered to finance and operate treatment
plants, waste diversions and collection facilities, reser-
voirs, and so on. Indeed, one of the purposes for studying
River Basin Authorities at this time, when so few are in
operation, is to clarify the proper role that each of these
powers and authorities must play.
The River Basin Authority has a formidable task. In order to
perform its function it has to alter the established and con-
venient practices of influential social units such as indus-
trial firms, cities, and even state governments, and to
encroach upon their former prerogatives. Its responsibilities
are defined in terms of a range of activities that overlaps
10
-------
and conflicts with the jurisdiction of older, established
governmental units. In such circumstances, the effectiveness
of the Authority in imposing and implementing its standards
will depend at least as much upon its ability to persuade as
upon its power to compel. The approach of this research,
therefore, has not been to analyze this governmental unit as
if it were in a position to set forth objectives and ordain
performance. Rather an attempt has been made to analyze the
Authority in the light of its problems of gaining adherence
to its goals and acquiescence to its regulations and charges.
A Conceptual Model of a River Basin Authority
The outcome of this investigation has been a new conceptual
model of the functioning of River Basin Authorities. The
model can utilize data and subjective judgments about tech-
nological, economic, and political conditions in a river
basin and can give indications of the type of quality manage-
ment plans that are likely to be adopted in the circumstances.
It is, therefore, a predictive theory of the performance of
this type of social institution.
The key concept underlying the model is that of the "Pareto-
admissible" pollution abatement plan. The primary role of
the River Basin Authority is seen as that of arbitrator
—balancing the partially conflicting desires of the indus-
tries, municipalities, conservation groups, governmental
agencies, and other interests that make up its widely varied
constituency. The Authority operates under a number of
technical, economic, and political constraints, and within
these limits it must find a solution that is attractive to
some if not most of the interested parties, and one that is
not absolutely unacceptable to any of them.
Since the Authority is sensitive to its constituents' interests,
it is safe to assume that it will be only by inadvertance
that an opportunity to aid any one interested group will be
passed up, provided no others object. Those decisions, or
quality management plans, that offer no opportunity to further
aid any one group except at the expense of others, are termed
"Pareto admissible." If, as Dorfman and Jacoby argue, it may
be presumed that the River Basin Authority's choice will be
restricted to the subset of Pareto-admissible pollution abate-
ment plans, then the range of likely choice is drastically
reduced. Where in the range of permissible choice the ultimate
decision will fall depends on the relative influence of the
contending parties within the decision process of the Authority.
Dorfman and Jacoby have shown how to construct a mathematical
problem whose solution will be one of the Pareto-admissible
decisions available to an Authority. They identify each of
11
-------
the relevant interest groups, 1, that may be concerned with
the choice among alternative quality management plans, X.
Depending on the decision.taken, each group realizes some
level of net benefits, NB (X). The mathematical problem is
constructed by assigning political weights, v^, to the various
groups concerned and then finding the decision that maximizes
the weighted sum of net benefits, denoted by ZwiNB^X), while
satisfying constraints and relationships that describe the
technological connection between abatement measures under-
taken and quality achieved, between abatement level and cost
Incurred, and so on.
The range of admissible choice may be traced out by assuming
many divergent sets of relative political weights, w±. The
ultimate outcome, naturally, depends on the influence actually
exercised in the circumstances—and it is not-possible to pre-
dict exactly what these relative weights may be. It is poss-
ible, however, to apply knowledge of the political facts of
life and judge some of the weight allocations as more likely
to be experienced than others. And based on these estimates,
one may make similar judgments regarding the decisions they
call for.
It is, perhaps, unnecessary to argue that a predictive theory,
such as the one proposed as a result of this research, may
have practical as well as scientific utility. The decisions
of any governmental body are predicated in part on expecta-
tions about the way in which other governmental bodies will
respond to them, just as the actions of the Federal Water
Quality Administration are conditioned by the expected
response of state and local water quality control agencies.
A theory that improves predictive ability can therefore
improve governmental decisions.
Besides, at the present time legislation for the establish-
ment of local water quality control authorities is in the
process of being drafted and tested. A predictive theory
will indicate, among other things, how various provisions of
enabling legislation and charters are likely to influence the
performance of local authorities. By so doing the theory
can assist in resolving some of the still moot questions about
the proper conditions and powers to confer upon quality con-
trol agencies.
A Preliminary Test of the Conceptual Model
In order to subject the conceptual model to a preliminary
test, it has been applied to a hypothetical river basin. The
analysis makes use of realistic technical and economic data;
the basin itself has been kept simple in order to facilitate
this preliminary application. Several municipal and industrial
12
-------
dischargers contribute to the generally poor quality of the
sample river; most of these same agents, along with others
both in and outside the basin, have an interest in bringing
the river to a higher quality standard. The question,
naturally, is how high a standard to strive for and how to
achieve it. The model focuses on a river basin pollution
control commission charged with formulating and implementing
a regional plan, and the analysis considers the way economic
and technical factors combine with political considerations
to influence commission decisions.5
Municipal Waste Management
In an attempt to cover some of the key aspects of the economics
of water quality, special attention has been devoted to the
municipality as a participant in the overall management pro-
cess. Most community water treatment facilities are organized
under the municipal governmental structure, and the success
of the Federal pollution control program depends in large
measure on the response of municipal authorities to national
policies.
Difficulties Faced by Municipal
Wastewater Managers
Of course, municipalities handle not only domestic sewage but
also the wastes of certain industries within their municipal
sewer service areas. For example, under current water pollu-
tion legislation, large Federal and state grants are available
to ease the burden of the construction cost of municipal waste
treatment facilities, but no cost sharing or incentive schemes
of any magnitude have been agreed upon to aid industries who
are managing their own wastes. There is, therefore, a strong
incentive for industries not previously served by municipal
systems to seek such a connection. This presents the municipal
water manager with the problem of allocating the financial
burden among industrial firms, both those already connected to
the sewer system and those joining in, and between business
^This experimental application, along with a summary of the
conceptual model, has been published in an article by Robert
Dorfman and Henry D. Jacoby entitled "A Model of Public
Decisions Illustrated by a Water Pollution Policy Problem."
The paper appeared first in a Congressional Joint Committee
Print [4], and subsequently a revised version has been pub-
lished in an edited volume on public expenditure and policy
analysis [5].
13
-------
firms and domestic waste sources. The mix of policy instru-
ments used by the municipal water and waste disposal system
manager to deal with these divergent interests influences
the effectiveness of the Federal program. In order to explore
this question, we have undertaken a study of what actually
happens when quality management decisions are faced by a
municipality serving large industrial polluters.
A Case Study of Municipal Wastewater Planning
Once again, as in the case of the studies of municipal water
systems under shortage, we were fortunate to encounter two
New England municipalities—Nashua, New Hampshire and Fitch-
burg, Massachusetts—that serve as excellent case examples.
The two are in the same river basin: Fitchburg is the major
upstream polluter of the Nashua River before it reaches the
town of Nashua, and both these cities contribute to the poor
quality of the Nashua as it merges with the Merrimac River,
Both are small municipalities with substantial industrial
sectors and both contain industries (tanning and paper)
which are heavy polluters. In addition, both towns have been
in the midst of planning new treatment facilities and in nego-
tiating the town response to recent Federal water pollution
legislation.
The report by Don Howarth on "The Federal Water Pollution
Control Program: A Case Study of Two New England Towns" [6]
presents the results of our study of the impact of the pollu-
tion control program on these municipalities. There are
several major issues to which this study is addressed. The
first concerns the alternative fiscal instruments available
to a municipality for financing treatment cost. The two
basic schemes which are available to municipalities are user
charges and property taxes on assessed evaluation. It is true
that if a charge system is chosen, it can be constructed in
many different ways. It appears in our studies, however, that
the critical municipal decision is whether or not to use user
charges at all; and an attempt is made to build a model of
this decision in order to analyze the influences which affect
it.
The Howarth report examines the issue of property taxes versus
user charges in terms of both efficiency and equity; moreover,
the report discusses the degree and direction of income dis-
tribution under each scheme, and it is argued that a charge
system is preferred on this score. In addition, it argues that
economic efficiency can only be achieved through the establish-
ment of the user charge alternative. The report goes on to
show, however, that when the choice between the two financing
schemes is left in the hands of local decision units, there
are circumstances under which the municipality can be expected
to choose the property tax alternative.
-------
In order to pursue the impact of different financing arrange-
ments on the municipality, the report examines the question of
whether the choice of financial instruments would be expected
to have an effect on the municipality's bond market position,
and vice versa. Finally, the report recommends changes in the
Federal water pollution control legislation which would guide
municipalities toward improved management policies from both
an efficiency and an equity standpoint. Specifically, the
report argues that some form of user charge scheme should be
made a prerequisite for Federal support of municipal treat-
ment plant construction.
Municipal Evaluation of Alternative Quality
Management Plans
A study by Clifford S. Russell provides a closer inspection
of the empirical problems of implementing the conceptual model
proposed by Dorfman and Jacoby. The attitudes of municipal
authorities are an important influence on the measures taken
by River Basin Authorities. As a result, the prediction of
likely basin-wide quality decisions requires an estimate of
the value of alternative plans—the NBi(X) term introduced
above—as viewed from the standpoint of participating cities.
The Viewpoint of Municipal Decisionmakers
In the planning of regional quality control, major munici-
palities will generally be represented by their elected
officials, who will be called upon to evaluate alternative
basin schemes, and who probably will have the power to commit
their respective cities to future changes in waste manage-
ment. In carrying out their responsibilities these officials
will, of course, be interested in the balance of costs and
benefits for their constituents that each of the alternative
proposals implies. A prime consideration in any negotiations
on a basin plan, therefore, will be the cost of new municipal
waste collection and treatment facilities and the distribution
of these costs among the city's residents. On the other side
of the balance will be a number of advantages: increased
recreational opportunities (at the city's waterfront or else-
where), an improved aesthetic situation in the vicinity of
the riverbank or elsewhere, and higher quality water at the
intakes of the municipal water supply system or of local self-
supplied industries. Judgments about these costs and benefits,
and their distribution, will determine whether a city will
support or oppose any particular abatement plan.
Three questions form the basis of Russell's study of this
phenomenon: (1) what are the categories of benefits and costs
that deserve to be considered, (2) what are the methods by
15
-------
which the magnitudes of these effects may be estimated, and
(3) what is the process by which we may visualize these con-
siderations being combined to form the overall judgment of
the "value" to the city of a particular abatement plan?
A Model of Municipal Decision Processes
The results of the research that follows from these questions
are reported in a paper that forms one of the chapters in the
new water pollution monograph. Russell first develops a des-
criptive model of the process by which municipal decision-
makers balance the costs and benefits that a particular
scheme implies for different groups of constituents. The
model shows how municipal officials may be imagined to go
about integrating these data to form their decisions to
support or oppose particular basin-wide plans.
The Russell approach to the functioning of governmental insti-
tutions is an extension of that proposed and tested in the work
of Dorfman and Jacoby. The government is viewed as having as
one of its chief preoccupations the maintenance of some level
of support among the set of local interest groups; it is very
important to a city administration how the costs and benefits
are distributed among these groups. Part of the problem, of
course, is that different groups carry different "weights" in
the minds of political leaders; some will be more politically
influential, some more "deserving" on one basis or another.
Together the influence and deservingness of a group will
determine the relative amount of attention that the leadership
feels must be accorded it in assessing the overall impact of
a particular program.
Once again interest groups are defined—only now the interests
are internal to the city. Individual groups may be indicated
by the index J. The gross quality benefits a group realizes
from a particular abatement plan X are determined by the
quality improvement q,, that the plan yields. Russell defines
the marginal benefit to group j as iJ>.(qY).
J A
If the total annual costs of compliance with the plan that
would bring quality change q,, are C, and the share of
group j, under a particular financing scheme is y., then the
annual net benefits of the plan to the group are: ^
NB1
16
-------
Next, Russell assumes that political support for the city
administration on the part of a group will be positively
related to the size of the group's net benefits resulting
from the river quality change and accompanying costs.
Further, the city administration may be viewed as having as
one of its principal, if not its overriding, goals the main-
tenance of support among its constituents. Since there will
be, at any one time, only a relatively small number of key
issues on which decisions must be made (and on the basis of
which political support can be gained or lost), and since
the significant interest groups will be known to politicians,
probably no great violence .is done to reality by assuming
that these city officials act in making the decisions as
though they perceived some (possibly weighted) set of lower
limits to the levels of net benefits which may be allowed
to accrue to each of the groups. Such a set of constraints
may be written formally as NBJ jl b.
Two possible sources for these values b* are considered.
One alternative is that these values are determined by the
local political decisionmakers themselves. If this is the
case, we may argue that the set of bj already reflects the
relative political importance of the several groups. If, on
the other hand, the bj represent legal constraints resulting
from previous local legislation or from legislation at the
state or Federal level, then the question of reflecting local
political weights is still open. For example, if group k
is a public utility, and if state legislation requires that
public utilities be permitted to earn a certain "fair return
on invested capital," bk would be selected so that when
NBk = bk, the utility's profit rate would be at the minimum
permissible level. But if the utility is politically power-
ful, it is likely that the local decisionmakers would feel
under pressure to devise plans for which the utility's actual
net benefit was larger than this implied legal minimum (or its
loss less than the legal maximum). In this situation one"
might simply imagine that the local authorities choose a new
set of bj such that all outside constraints are at least
satisfied and that the more powerful local groups are, in
fact, assured of results superior to the minimum levels
required by law. Or one might use a model employing weighted
net benefits, as does the basin-wide model described earlier.
Russell explores all these approaches.
The Importance of Cost-Sharlng^Schemes
Within the Municipality
For a given q,, and cost-sharing scheme, therefore, there
will be a value of C that is the largest cost that can be
shared without violating a perceived political constraint.
17
-------
(This value may be zero.) Of course, manipulation of the
vector y. ™ay permit a higher cost which is "sharable"
within the constraint set. Thus, for a given qx and a
particular form of the constraint set, changing y. in such
a way as to redistribute the cost burden toward those groups
enjoying large benefits from q^ will tend to increase the
level of C at which the constraint becomes binding.
There will, in general, be fundamental institutional con-
straints on the vector y ; for example, cities may be con-
stitutionally unable to introduce sales or income taxes, to
change water rates or to charge for liquid waste treatment
on a volume and waste concentration basis. Available tax
and charge options may only allow the variation of y, over
a small region of the relevant space.
In these terms, then, the city's willingness to pay for a given
quality change may be defined as the highest level of cost, C,
which can be accepted by the city government in complying with
that basin plan, subject to the institutional constraints on
the variation of the y^'s and to the political constraints
on the resulting net benefits accruing to the interest groups.
A Sample Application of the Model
This conceptual model, and the empirical data developed,
have been subjected to a test application. The interest
groups that were identified include the domestic sector,
tourism-based industries, heavy polluting industries, and
commercial and industrial establishments not otherwise
classified. The "worth" to these groups of quality improve-
ment includes benefits from
—aesthetic betterment and increased
recreational opportunities related
to the river,
—lower water supply treatment costs
due to improved upstream treatment
of oxygen-demanding wastes, or
—lower water supply treatment costs
due to lower levels of dissolved
solids in the stream.
The sample case is built on the earlier work by Howarth [6]
and involves the city of Nashua, New Hampshire.
18
-------
Estuarine Water Quality Management
During the period from 1961 to 1967 an extensive survey and
analysis of pollution in the Delaware Estuary was conducted
by the-Delaware Estuary Comprehensive Study (DECS) of the
Federal Water Quality Administration. Since we are using
the Delaware River Basin and Estuary as a laboratory example
in much of our research, we have been fortunate to have the
close cooperation of the members of the Delaware Estuary
Comprehensive Study group as well as of Dr. Robert V.
Thomann who directed the original work. Many of the data in
our studies were provided by the DECS team.
As we began our work on the economics of water quality, it
was. necessary for us to pull together the existing informa-
tion on the Delaware Estuary conditions and to gain familiarity
with the systems analysis models which had already been
developed for that situation. The study by Grant W. Schaumburg,
Jr.> "Water Pollution Control in the Delaware Estuary" [17],
was our initial effort to develop this capacity and to extend
the analysis and add our own contribution to it.
The Schaumburg study uses linear programming models of estuarine
pollution abatement to analyze the economic characteristics of
different quality management policies. Among the different
policy measures considered are the following.
(1) "Required treatment" schemes whereby polluters
are required to subject all wastes to a stan-
dard treatment process before discharging them
into the watercourse.
(2) "Effluent standard" measures whereby the regu-
latory body specifies how much waste can be
discharged by each polluter or by how much
each polluter must cut back his waste
contribution.
(3) "Effluent charges" whereby polluters are taxed
according to the quantity of pollutant discharged.
In addition, the study considers the use of a variety of zoning
procedures in combination with these different control measures.
Zoning consists of dividing the river or estuary into several
different sections or zones according to the type of economic
activities which will be allowed or the kind of treatment
facilities which must be provided. Zoning is designed to
encourage a degree of specialization in stream use and is
an aid in seeking abatement solutions that do not require
heavy expenditure by polluters who are only making a marginal
contribution to the zones of maximum degradation.
19
-------
The Schaumburg paper presents a clear exposition of the linear
difference equation model of estuary dynamics, which is the
basis of much current work on estuarine quality improvement.
It also provides a clear and concise development of the linear
programming models used to analyze the technical and economic
characteristics of the various policy measures discussed
above, and it applies these techniques to an analysis of the
Delaware Estuary
Research on Computing Techniques for Basin-wide Planning
Interaction of Screening Models and Basin Simulation
One of the key aspects of our research over the grant period
has been an effort to develop improved systems analysis models
for basin-wide planning of water resources. This work has been
based on the use of digital simulation techniques. The simu-
lation model with which we have been -working is the product
of an earlier phase of the Harvard Water Program research,
and it encompasses the Delaware River Basin above Trenton,
New Jersey along with the Schuylkill River upstream from
Philadelphia, Pennsylvania. In its current version, the
model allows consideration of thirty-five existing and
potential reservoir sites and takes account of water use for
recreation, municipal and industrial water supply, electric
power generation, flood control and the maintenance of instream
flow for quality control.
The model is designed to permit analysis of the interactions
among water uses in all the sub-basins of the overall
Delaware system. In studying the economics of water supply
and quality, it is necessary to consider the Delaware River
Basin as a whole, because, for instance, the need to supply
water to the city of New York from the upper part of the
Basin may conflict with the need to supply water to combat
saline pollution in the Delaware Estuary downstream. A
simulation model of such a large multipurpose, multi-unit
system, however, is necessarily large and complex, and con-
siderable effort has been devoted to the maintenance and
improvement of this particular computer program.
The Need for Preliminary Screening Techniques
In our work with this model, we have encountered several
difficulties which have led us to shift the emphasis of our
research on computation methods for use in planning water «
systems. There are two major problems that arise in the
application of large river basin simulation models: one is
the selection of physical designs from the many possible con-
figurations of system components, and the other is the formu-
lation and "tuning" of a reservoir operating policy for the
20
-------
particular system under study. In an attempt to alleviate
these two problems, we have turned our attention to the use
of mathematical programming models for preliminary screening
of system designs and for an initial formulation of system
operating rules.
In order to emphasize the potential usefulness of an effec-
tive preliminary screening model, it is helpful to review
briefly the way basin simulations are applied in practice.
The analyst must select a set of physical plans or "designs"
for study, where each design consists of a set of reservoir
sizes, construction schedules and target outputs for different
water services. Then, using a predetermined reservoir release
policy, the model is used to simulate the operation of the
system over some period of years and to compu-te indicators of
the economic and physical characteristics Of system perfor-
mance. Even in a system with only three or four potential
reservoirs, the number of potential designs is very large,
and considerable engineering judgment usually goes into the
selection of plans for detailed analysis by simulation.
In some studies, statistical sampling techniques have been
used in the selection of alternative river basin plans. In
an earlier version of our Delaware simulation model, which
was applied to the Lehigh River (a sub-basin of the Delaware)
[7], for example, the design variables included the sizes
and t'he timing of reservoir construction, target levels of
power output, target levels of reservoir storage for recrea-
tion, target M & I and water supplies, allocation of flood
storage and desired low-flow controls, etc. Random samples
were taken of the individual variables in sequence in order
to generate a set of alternative designs; minor adjustments
were then made to each of these sample designs in search of
the "best" design for the system as a whole. In the applica-
tion to the six-reservoir system of the Lehigh River, thj-c
approach proved quite successful.
In the case of the overall Delaware Basin simulation, which
has many more design alternatives, the sampling approach
appears far too time consuming and expensive. Indeed, the
system offers so many choice variables that it becomes
difficult to find a basis in engineering judgment for a set
of trial designs; and even given a trial design it is diffi-
cult to formulate marginal changes which improve the overall
economic performance.
Mathematical Models for Preliminary Screening
In an attempt to find ways to simplify this task we took up
the study of simpler, faster and more flexible basin-wide
21
-------
models which—although less accurate than the simulation
approach—could be used to screen out a large proportion of
the possible designs as economically inferior and to gain a
rough indication of which system configurations might turn
out to be the more attractive. In the pursuance of this
work, we were fortunate to have the cooperation of Daniel P.
Loucks of Cornell University. During the spring and summer
of 1968 he was associated with the Water Resources Group in
the Division of Engineering and Applied Physics at Harvard —
his primary effort being devoted to the study of mathematical
programming models of water resource systems. His work with
the Division proved to have a productive application in con-
nection with our problem of preliminary screening; discussion
of the structure of the mathematical models tailored to this
application is provided in a paper by Loucks, "Preliminary
Screening Models for River Basin Planning" [8].
In approaching the study of preliminary screening models for
use in combination with digital simulation, it was decided to
start not with the overall Delaware Basin but with a smaller
and more tractable system; and the first efforts were devoted
to screening-simulation studies of the Schuylkill River Basin.
This is a five-reservoir system involving municipal water
supply, recreation, flow augmentation, and a potential for
power production. A series of different types of mathe-
matical programming models was developed for this system,
and the Delaware Basin simulation was revised to allow con-
sideration of the Schuylkill as a separate entity. The simu-
lation program was run in concert with the screening models
in order to check their performance in representing system
characteristics and to investigate alternative ways of using
the information produced by the mathematical programming
models.
The results of this work have been encouraging: experience
to date indicates that these preliminary screening models can
be set up and solved with a reasonable input of time and
expense, and that they can be a great help in identifying
those system plans that are worth studying in detail by
digital simulation.
Systems Models for Evaluation of Flow Augmentation
The larger the flow in a river or stream the more waste
material it can absorb without serious ecological change
and consequent loss in recreational and aesthetic value.
Because this is so, acute water pollution problems occur
most often during periods of low streamflow, and the require-
ments of the low-flow period usually are a controlling factor
in the design of waste management systems. Low-flow condi-
tions can be improved at a cost, of course, and in some cases
22
-------
investment in flow regulation may be an effective and efficient
component of quality control policy. The purpose of research
in this area has been to develop analytical models that des-
cribe the effects of flow regulation, and that can be used
to evaluate the relative attractiveness of reservoir storage
when used in concert with other quality control instruments.
The analysis of flow regulation schemes turns out to be a
rather complex task, for several reasons. Consideration must
be given to stream hydrology, to reservoir construction, and
to the combination of storage system design and operation
that will achieve specified degrees of flow regulation. The
evaluation is made more complicated by the fact that regulating
reservoirs often serve multiple purposes, such as recreation,
water supply, flood control, or power generation. And, to
add to these difficulties, careful analysis is required to
define the interaction between streamflow control and waste
treatment as they affect water quality at different points
downstream.
As mentioned in the previous section, the analysis of multi-
purpose reservoir systems can be a complicated and expensive
matter. Often, the analysis is based upon digital simulation,
and to get the most out of expenditure on these simulation
studies there is need for a procedure that can be used to
screen large numbers of possible flow regulation schemes and
pick out the relatively small range of alternatives that
merits more thorough investigation. Once again, it is toward
this preliminary screening problem that the model development
efforts have been directed.
Consideration of the Political Dimension
In eliminating or "screening out" the less attractive oppor-
tunities, consideration is taken of political as well as
technical and economic factors. The term "preliminary
screening" normally refers to the rapid, crude analysis of
projects in search of a small number that are technically
feasible and that appear to perform well under some esta-
blished standard of evaluation or objective function. This
research begins from a different viewpoint, and argues that
the precise statement of relevant objectives that is required
for mathematical analysis quite often is not available, par-
ticularly in the early stages of project, evaluation. At best,
objectives are worked out as the project develops, and in
many instances the purposes for which a project is undertaken,
and the influences that shape its design, may never be formu-
lated explicitly. This occurs because, in a public decision
such as quality control by flow augmentation, there will be
several groups—Federal and local, private and public—whose
23
-------
interests in the outcome may diverge, and yet who carry sig-
nificant weight in the ultimate choice.
In this circumstance, any preliminary screening process, if
it is to be useful, must incorporate an extra consideration
besides pure technical and economic feasibility; it must have
an eye to the potential for political survival of alternative
proposals. It does little good to identify projects that look
attractive in the light of the objectives of some particular
group, but that will never get beyond the drawing board because
of factors the analysis did not incorporate. The successful
analyst carries these considerations in the back of hi-s head,
of course, and tries to take them into account when and where
appropriate. By building our analysis on the conceptual frame-
work laid down in the work on political factors summarized
above, however, it becomes possible to make explicit provi-
sion for the political dimension of the screening task.
Mathematical Models for Basin Analysis
Much of the development and testing of flow augmentation models
has been carried out using a case example. For this purpose,
the hypothetical river basin utilized by Dorfman and Jacoby
[3,5] serves well, particularly since use is made of the model
of political decision which that basin was designed to elaborate,
Most of the descriptive information developed for the former
study is carried over for use in the analysis of flow regula-
tion. Of course, the basin has been expanded to include flow
regulation as an alternative quality control instrument.
Several mathematical models are used in the overall analysis.
One set of models is designed to analyze multipurpose reser-
voir projects. A reservoir release rule has been developed
that is compatibele with the low-f.low conditions specifier!
in most water quality standards, yet is suitable for inclu-
sion in analytical models. Applying this release procedure,
a mathematical model of a multipurpose reservoir system has
been formulated, using recreation and flow regulation as com-
panion purposes for the storage facilities. Because reservoir
pool elevation affects recreation activity, the model has been
designed to permit consideration of fluctuations in storage
that are imposed by operation for flow control. Of course,
i-ecreation is only one of many reservoir uses that is
affected by variation in water level; it
-------
to this set of models, it proves possible to screen out
large regions of possible system designs and focus attention
on a relatively small set of alternatives that appears to
merit further study.
The results of this research are reported in an essay by
Daniel P. Loucks and Henry D. Jacoby [9].
Research on the Use of Regional Analysis in Water
Resource Planning
Regional Analysis of Quality Control Costs
In the original formulation of this research project, a col-
laborative effort was envisioned whereby we would work along-
side individuals in the Department of Regional Science of the
University of Pennsylvania and the Regional Science Research
Institute in an attempt to integrate,methods of regional eco-
nomic analysis with studies of the economics of water supply
and quality. The primary emphasis of the work of the regional
science group has been on the potential use of regional input-
output techniques to estimate future demands for water supply
and water disposal in the region of the Delaware Estuary and
to devise methods for forecasting the spatial distribution of
waste, loads along the estuary. We have maintained close work-
ing contact with their efforts over the grant period.
In addition to our work in cooperation with the regional
science group, we have undertaken our own study of the appli-
cation of regional input-output methods to water quality
problems. A paper by William E. Martin, "The Use and Value
of Input-Output Models in the Analysis of Water Pollution
Abatement" [11] explores potential applications of this tech-
nique. The paper begins with a brief description of the basic
components of input-output models and of the way in which such
models may be extended to include data related to the use of
primary resources. The discussion focuses on data relative to
water quantities and qualities. A second section presents a
general discussion of the possible value of input-output
analysis of problems of pollution abatement, and the final
section develops an empirical example of one possible use.
The case example, once again, is the Delaware Estuary, and an
input-output model is used to estimate the relative sizes of
abatement costs—and on whom they would fall—if pollution
were controlled by treatment plants, or alternatively, by
restrictions on industrial output. Although designed
primarily as an example of the way in which regional analysis
might be applied to pollution problems, the study also con-
cludes that, in the Delaware case, treatment costs are
25
-------
relatively small compared to the direct and indirect effects
on the output of industries such as petroleum, chemicals and
paper, which are large contributors to the regional waste load.
A Model of the Interdependence between a Regional
Economy and Water Quality Control Efforts
A second, similar research effort has been devoted to the
preparation of a model that describes the interdependence of
a regional economy and a river flowing through it. What is
desired is a model that can be used to measure the effect of
regional growth (and associated locational changes) on a
river's reaches and to assess the influence of alternative
regional waste management policies. And this orientation
leads to a concern with the way expenditures in the water
supply and sewage treatment sectors tend to'feed back into
the regional economy—resulting in increased levels of out-
put and employment and, unfortunately, increased levels of
pollution generated by the region's industries. The analysis
of these phenomena rests on a static regional input-output
model that has been expanded to include consideration of the
river, and of the geographical distribution of waste generat-
ing activities.
Under this mode of analysis, all water users and waste sources
in a region are considered. The usual distinctions are made
between intermediate industries, households, and government
of all levels operating in the region. And, for purposes of
analyzing water supply and quality, these entities are further
classified into three groups of water users. The first group
consists of units that rely on water supply and sewer services
for their intake and discharge of the river's water. The
second includes those who do not use public water and sewer
services, but are directly linked to the river by withdrawing
water and subsequently discharging it after some level of
treatment. The third group contains water users who either
(a) do not use the river's water, but rely on groundwater
sources and then discharge into the river, or (b) are located
in another drainage basin and use the water supply and waste
disposal services of some other river.
The key linkage between these diverse elements of the river
and the regional economy, then, is provided by the water supply
and sewer services that transport water and alter its quality
through various treatment processes. There are well-defined
areas within a metropolitan region that are linked to the
river at specific points of water intake and sewer outfall.
And, therefore, in order to measure the locational incidence
of wastewater discharge, the spatial distribution of water
users, as they are tied to different reaches of the river,
must be specified.
26
-------
In conventional input-output analysis, however, no information
is available regarding activity levels by subarea, or about
flows of goods and services among subareas of a region. Con-
sequently, to carry forward this regional analysis it is neces-
sary to partition the metropolitan region into local areas
corresponding to water supply and sewer service systems. The
development of techniques for spatial disaggregation of water
utilization coefficients (intake, discharge, wasteload) has
been one of the major components of this research effort.
Once the model has been formulated, it becomes possible to
simulate the growth of a metropolitan region over time and
to assess the impact of the regional economy on the river
under alternate policies regarding level of waste treatment,
flow regulation, regionalized treatment, etc. The spatial
dimension of the growth of wasteloads thus may be considered
in the analysis of alternate management schemes, and it
becomes possible to evaluate the significance of alternative
control schemes on the regional economy itself.
Preliminary results of this research are reported in a paper
by Eliahu Romanoff, "The Interdependence of a Regional
Economy and a River" [13]. Work on this topic has not been
completed, however.
27
-------
SECTION III
A MONOGRAPH ON WATER QUALITY MANAGEMENT
As mentioned at the outset, one of the purposes of Grant No.
16090 DTP has been to make possible the integration of indi-
vidual research efforts, along with the output of other water
quality investigations at Harvard, into a cohesive monograph.
An important goal of our research in water resources over the
years has been to bring together the knowledge of several
disciplines—physical and biological science, engineering,
economics, and political science—and to develop a coordinated
approach to water resource management. Twice in the past,
water research at Harvard has been drawn together in this
fashion—first in the book Design of Water-Resource Systems
by Maass, et al. [10],, and later in the monograph on Simula-
tion Techniques for River Basin Planning, by Hufschmidt and
Fiering L1J •' The preparation of these volumes not only pro-
vided a healthy exercise for the researchers and authors
themselves—for it forced them to bring their several disci-
plines to bear on common problems—but it also made research
results available to a wider audience of water resource
managers, analysts, and researchers than would otherwise have
been possible.
The latest monograph is a collection of nine essays selected
from recent research at Harvard. The book has been titled
Models for Managing Regional Water Quality.
The first essay, by Harold A. Thomas, Jr., provides a broad
perspective of the water pollution problem in an historical
setting. It reviews the antecedents of current practice in
water quality control and lays the groundwork of ecological
principles and engineering knowledge upon which the analyses
in subsequent chapters are built. Special attention is given
to the problem of biochemical pollution from human and indus-
trial wastes, and to the empirical basis of the mathematical
models commonly used to describe this phenomenon.
29
-------
The next three chapters present research (conducted under
this grant) by Robert Dorfman, Henry D, Jacoby, and Clifford
S. Russell. These chapters endeavor to integrate social and
economic considerations with the technological and engineering
aspects of water quality management, Chapter 2 by Robert
Dorfman, "Conceptual Model of a Regional Water Quality
Authority" [2], introduces the political model described
earlier and advocates its application to the analysis of
pollution control institutions and projects. In Chapter 3,
"An Illustrative Model of River Basin Pollution Control"
[3], Robert Dorfman and Henry D. Jacoby present the results
of the preliminary test of the model, using the hypothetical
river basin, And in Chapter 4, "Municipal Evaluation of
Regional Water Quality Management Proposals" '[11] » the results
of the Russell study discussed above are detailed.
Chapters 5 and 6 present the work by Stephen J. Turnovsky.
The first of these chapters, "Microeconomic Behavior When
Supplies Are Uncertain" [25], explores^ the behavior of micro-
economic units—consuming households "and productive firms—
when the supply of some input is subject to uncertainty.
Applying and extending these results, a second chapter,
"Optimum Government Investment in Supply Systems Yielding
Uncertain Outputs" [26], considers the optimal planning of
public investment and management schemes when there is a
stochastic component in the pattern of goods they provide.
Particular attention is devoted to the estimation of benefit
functions in this situation, and to the design of supply
systems when output variance is subject to control.
The last three chapters in the book investigate some of the
different combinations of control instruments that may be
applied to particular regional water quality problems. Waste
reduction at the source is the most common method of pollution
abatement, but there are others. For example, it is possible
to apply remedial measures to the water body itself, through
the use of flow augmentation, artificial aeration, or the
removal of blankets of sludge or coatings of oil. Alterna-
tively, wastes can be collected from diverse sources and
treated in a central facility, or they can be piped out of the
region altogether,
Generally, the efficient use of such measures requires close
coordination with plans for waste treatment by individual
dischargers, for they can change the magn-itude and geographi-
cal distribution of the assimilative capacity of the receiving
waters, or of the waste loads themselves. Study of these
alternatives, therefore, is best carried out in the context of
systems models that can encompass a complicated interdependence
among the choices that are open. Each of the last three
studies is a quest for new capacity t-o perform this type of
analysis.
30
-------
Chapter 7 by Leonard Ortolano, "Artificial Aeration as a
Substitute for Wastewater Treatment" [12], is concerned with
the use of artificial aerators for control of dissolved oxy-
gen in estuaries. In a sample problem, which incorporates
data from the Delaware Estuary, a study is made of the choice
between the introduction of mechanical aerators and the instal-
lation of advanced waste treatment, Several aspects of such
a system render this a challenging analytical problem. The
relationship between aerator performance and ambient stream
conditions is a complicated one: the lower the dissolved
oxygen level, the greater the rate of aeration from an
artificial device, Furthermore, aerators can be installed
at any point in an estuary, and they can be easily removed
or shifted to other locations as conditions change. Advanced
waste treatment facilities, on the other hand, are fixed in
place, but they are subject to economies of scale in plant
size, Capacity tends to be built ahead of demand, as a
result, and the planning analysis must cover multiple time
periods. These characteristics contribute computational
difficulties, and the bulk of the paper is devoted to the
development of mathematical programming formulations and
computer algorithms to deal with them,
Chapter 8 by Tze-wen Chi, "Wastewater Conveyance Models,"
[1], considers wastewater conveyance and regional treatment
facilities, The analysis is similar to that applied to
aeration. The sample problem is concerned with multiperiod
planning of estuarine quality control, once again using data
from the Delaware, In this chapter, however, the alternative
to on-site waste removal is to transport wastes to some com-
bination of regional treatment plants, or to pipe them out of
the region.
Not surprisingly, the consideration of pipe design introduces
special analytical difficulties of its own. There are econo-
mies of scale in pipe construction; and the relationships
among pipe diameter, flow rate, pump design, and energy cost,
though well understood, are complicated. In addition, it can
prove efficient to install a trunk sewer that flows one way
in one time period and in the reverse direction the next.
Thus the analysis must consider pipe capacity, flow rate and
direction, as well as the complexities of treatment plant
design and operation, As in the preceding chapter, much of
the discussion is directed to mathematical and computational
procedures adequate to the problem,
Both the Ortolano and Chi studies were carried out under the
sponsorship of PWQA grants other than the one being reported
upon here. They were assisted by grants on "Operations
Research in Water Quality Management" (Grant No. WP-00111) and
"Dynamic Planning Techniques for Pollution Abatement" (Grant No.
16090 DSS).
31
-------
The chapters on aeration and wastewater conveyance give full
attention to the solution of analytical problems inherent in
the economic and technical characteristics of those control
measures. They do not emphasize the institutional orienta-
tion of the early chapters. Chapter 9 on "Flow Regulation
for Water Quality Management" [9], by Daniel P, Loucks and
Henry D, Jacoby, reports on the study of flow augmentation
that was described earlier, and this last essay does inte-
grate the technical work on quality models and the institu-
tional concerns of the early chapters.
That completes the book. As this brief summary indicates,
most of the work is exploratory, Furthermore, the emphasis
throughout is on the development of analytical models that
can encompass the technical and social facets of a compli-
cated subject, and the coverage is necessarily spotty.
There is much discussion of biochemical pollution but scant
attention to other critical problems such as waste heat,
radioactive discharge, inorganic industrial chemicals, and
agricultural runoff, There is a good deal of work on models
suited to rivers and .estuaries; little effort is devoted to
lakes, bays, and coastal areas, Municipalities are subjected
to close scrutiny; independent industrial dischargers are
handled in summary fashion.
Thus, in a sense, these studies comprise less a book about
water quality than one about methods for analyzing problems
of water quality. The research has been less comprehensive
in considering all instances of water pollution and more
integrative in attempting to tie together the physical,
economic, and institutional aspects of regional management.
Our hope is that these essays can provide a stepping-stone to
further research and to improved analysis for the hard,
practical task of environmental protection.
32
-------
SECTION IV
ACKNOWLEDGMENTS
Thanks are due to many people for their efforts in support of
this work over the past several years, The papers presented
here are drawn from recent research by participants in the
Harvard Water Program: these studies represent the latest
stage in a fifteen-year history of inter-disciplinary research
in water resources under the Program's auspices.
Since its inception! the Water Program has been organized
within Harvard's John F. Kennedy School of Government, and it
has worked hand-in-glove with the water resources group in the
Division of Engineering and Applied Physics. The research
activity has enjoyed strong support as well from the Depart-
ment of Economics, It is difficult to imagine better circum-
stances for this type of cooperative effort, and our debt to
the leadership of these departments is great,
The primary vehicle for organizing and sustaining the research
effort has been a graduate seminar sponsored jointly by the
Economics Department and the Kennedy School, From semester to
semester its function has been the same, i.e., to provide a
forum for the presentation of ideas and preliminary drafts,
and an opportunity to receive criticism and suggestions. Each
of the nine chapters was presented to the seminar at one stage
or another: most were worked over more than once on the long
path from initial conception to completed papers,
What this means, of course, is that a good deal of hard work,
and often some very important suggestions-, were contributed
by persons who are not listed here. Every year the seminar
enjoys the active participation of several Harvard faculty
members, and of students from both Harvard and MIT. There
are too many to credit them all. There are several whose
omission would be a gross injustice, however, and we are happy
33
-------
to give special thanks to Robert Burden, Russell deLucia,
Joseph Harrington, Daniel Leucke, Arthur Maass, Peter
Rogers, Grant W. Schaumburg, Jr., and Richard Woodward.
Finally, much of this research might never havebeen under-
taken without continuing support, both financial and
personal, from Resources for the Future, Inc.
-------
SECTION V
REFERENCES
[1] Chi, Tze-wen, "Wastewater Conveyance Models," Chapter 8
(90 pp.) in Models for Managing Regional Water Quality,
R, Dorfman, H, D. Jacoby, H. A. Thomas, Jr., eds.
Being considered for publication, February, 1971.
[2] Dorfman, R., "Conceptual Model of a Regional Water
Quality Authority," Chapter 2 (90 pp.) in Models for
Managing Regional Water Quality, R. Dorfman, H. D, Jacoby,
H. A. Thomas, Jr., eds. Being considered for publication,
February, 1971.
[3] Dorfman, R,, and H. D, Jacoby, "An Illustrative Model of
River Basin Pollution Control," Chapter 3 (100 pp.) in
Models for Managing Regional Water Quality, R. Dorfman,
H. D, Jacoby, H, A. Thomas, Jr., eds. Being considered
for publication, February, 1971.
[4] ., "A Model of Public Decisions Illustra-
ted by a Water Pollution Policy Problem" in The Analysis
and Evaluation of Public Expenditures; The PPB System
(a Congressional Joint Committee Print), Volume T,Part II.
Washington: U. S, Government Printing Office, 1969.
[5] ., "A Model of Public Decisions Illustra-
ted by a Water Pollution Policy Problem" in Public Expend-
itures and Policy^Analysis, Robert H. Haveman and Julius
Margolis, ~eds~. Chicago: Markham Publishing Company,
1970.
[6] Howarth, D,, The Federal Water Pollution Control Program:
A Case Study of Two New England Towns, Harvard Water
Program Di scussion Paper No, 6b-3> T37 PP«i Harvard
University, 1968.
35
-------
[7] Hufschmidt, M,, and M. B Fiering, Simulation Techniques
for Design of Water-Resource Systems. Cambridge,
M a sVa c hu s ettf s~iHarvard University Press, 1966.
[8] Loucks, D, P., Preliminary Screening Models for River
Basin Planning,'Harvard Water Program Discussion Paper
No. 68-7, 30 pp. Harvard. University, 1968,
[91 Loucks, D. P., and H. D. Jacoby, "Plow Regulation for
Water Quality Management," Chapter 9 (120 pp.) in
Models for Managing Regional Water Quality, R. Dorfman,
H, D. Jacoby, H. A. Thomas, Jr., eds.Being considered
for publication, February, 1971«
[10] Maass, A., et al., Design of Water-Resource Systems.
Cambridge, Massachusetts: Harvard University Press,
1962.
[11] Martin, W, E,, The Use and Value, of Input-Output Models
in the Analysis of Water Pollution Abatement / Harvard
Water Program Di s cuss ion Pap e r No. 68-5, 32 pp. Harvard
University, 1968.
[12] Ortolano, L,, "Artificial Aeration as a Substitute for
Wastewater Treatment," Chapter 7 (100 pp.) in Models
for Managing Regional Water Quality, h. Dorfman,
H, D, Jacoby, H. A, Thomas, Jr., eds. Being considered
for publication, February, 1971t
[13] Romanoff, E,, "On the Interdependence of a Regional
Economy and a River," paper prepared for the Fifth
International Conferen-ce of Input-Output Techniques,
Geneva, Switzerland, January 1971, 20 pp.
[I1!] Russell, C. S., "Municipal Evaluation of Regional Water
Quality Management Proposals," Chapter ^ (130 pp.) in
Models for Managing Regional Water Quality, R. Dorfman,
H. D. Jacoby, H, A. Thomas, Jr., eds. Being considered
for publication, February, 1971.
[15] • i The Recent Northeast Drought! Short-
run Costs of Water Shortage and Lessons for Water System
Planning;', Harvard Water Program Discussion Paper No. bb-
10, 346,pp, Harvard University, 1968.
[16] Russell, C. S,, D. G, Arey, and R. W. Kates, Drought and
Water Supply; Implications of the Massachusetts Experi-
ence for Municipal Planning. Baltimore;Johns Hopkins
Press, 1970.
36
-------
[17] Schaumburg, G. W,, Jr. Water Pollution Control in the
Delaware Estuary, Harvard Water Program Discussion
Paper No, 67-2, 150 pp. Harvard University, 1967.
£18] Thomas, H, A,, Jr., "Waste Disposal in Natural Water-
ways," Chapter 1 (65 pp.) in Models for Managing
Regional Water Quality,• R, Dorfman, H, D, Tacoby,
H, A, Thomas, Jr., eds, Being considered for publica-
tion, February, 1971.
[19] Turnovsky, S, J,, "A Model of Consumer Behavior Under
Conditions of Uncertainty in Supply," International
Economic Review, to be published.
[20] ,, "The Behavior of a Competitive Firm
with Uncertainty in Factor Markets," New Zealand
E c onom i c Pap e r s, 3., No, 1 (November 1969).
[21] . i Stochastic Demand and the Theory of
the" j^lrm. Harvard Water Program Discussion Paper No,
6 «-2, 22 pp. Harvard University, 1968.
[22] ., Some Criteria for the Optimal Selec-
tion of Risky~Tnvestrnent Proj ect s, Harvard V/ater Program
Discussion Paper No,68-6,32 pp, Harvard University,
1968.
[23] • i The Theory of Production Under Condi-
tions of Stochastic Input Supply, Harvard Water Program
Discussion Paper No, 68-JJ, 26 pp. Harvard University,
1968.
[24] ., "The Demand for Water; Some Empiri-
cal Evidence on Consumers' Response to a Commodity
Uncertain in Supply," Water Resource Research, 5, No. 2
(April 1969), PP. 350-3ST;"
[25] ., "Microeconomic Behavior When Supplies
Are Uncertain," Chapter 5 (65 pp.) in Models for Mana-
ging Regional Water Quality, R. Dorfman, H. D.. Jacoby,
H. A, Thomas, Jr., eds, Being considered for publica-
tion, February, 1971.
[26] ., "Optimum Government Investment in
Supply Systems Yielding Uncertain Outputs," Chapter 6
(70 pp.) in Models for Managing Regional Water Quality,
R, Dorfman, H, D. Jacoby, H, A, Thomas, Jr., eds.
Being considered for publication, February, 1971.
37
------- |