Chicago Lakefront Demonstration Project benefit-cost analysis

[From the U.S. Government Printing Office, www.gpo.gov]

cp
Coastal zone
Information
Center
W.P. COASTAL ZONE
INFORMATION CENTER

CHICAGO LAKEFRONT DEMONSTRATION PROJECT

BENEFIT-COST ANALYSIS

PREPARED

LEE G. ANDERSON

RUSSELL F.SETTLE

UNIVERSITY OF DELAWARE

ROY F WESTON. INC

ILLINOIS COASTAL ZONE MANAGEMENT PROGRAM
Marina CitY Office Building
300 N. State Street Room 1010
Chicago, Illinois 60610

HD HICAGO DEPARTMENT OF DEVELOPMENT
47.4 ILANDIC AND PLANNING
A53
1977_

LEWIS W. HILL
COMMISSIONER

COASTAL ZONE
INFORMATION CENTER

CHICAGO LAKEFRONT DEMONSTRATION PROJECT

BENEFIT-COST ANALYSIS

PREPARED

LEE G. ANDERSON

RUSSELL F. SETTLE

UNIVERSITY OF DELAWARE

ROY F WESTON: INC

The preparation of this document was
financed in part by the National
Oceanic and Atmospheric Administration,
U.S. Department of Commerce under the
Coastal Zone management Act of 1972.

CITY OF CHICAGO
DEPARTMENT OF DEVELOPMENT
MICHAEL A. BILANDIC AND PLANNING
MAYOR LEWIS W. HILL
COMMISSIONER.

TABLE.OF CONTENTS

Page No.

SECTION 1
INTRODUCTION 1

SECTION:2
BASIC STEPS,-FOR UNDERTAKING.BENEFIT-COST ANALYSIS 6
2.1- Basic Steps-Outline 6
2.2 Step 1 - Define the Project 6,
2.3 Step 2 -.List Project Benefits .8
2.4 Step 3 - List Project Costs 10
2.5 Step 4@- Determine Which Benefits and Costs
Can.be Measured Theoretically 10
Z.6 Step 5 - betermine,Which Benefits and Costs
@Can be:Measured 12
2.7@ Step 6 - Determine What,Will be Measured'and How, 12'
2.8 Step 7 - Demonstrate Effect of Nonmeasured Items 13
2.9- Step 8-.Select,Proper Method of'Displaying
Information 14

SECTION.3
PROBLEMS,-OF MEASURING BENEFITS 16
3-..l Overview 16
3.2 Demand Curve: Marginal and Nonmarginal
Changes.In Output 17
3.1 Restrictions@on,Demand Curves 19
3.3.1 Price.Ceilings 19
3.3.2 Price@Floors, 22
3.3.1 Output Controls 22
3.4 Other Measures of Benefits. 23
3.4.1 Monetization of Value of Materials
Deterioration 24
3.4.2 Resulting from.Exposure.to Pollutants 24
3.4.3 Monetization of Effects on Vegetation 24
Effects on.Residential Property Values 25
3' .4.5 Willingness to Pay Survey 26
3.4.6 [email protected] Site Demand as a
Function-of Travel Cost 27
3.4.7 Reduction.in Cost,of Avoidance 29
3.4.8 Reduction iniCost of Treatment 29
3.4.9 Damages Awarded As a Result of Litigation 30
3.4.10 Value of Time. 31
3.5 Comparison of Measurement Methods 32

SECTION 4
PROBLEMS OF MEASURING COST 35
4.1 General 3 5
4.2 Unemployment
4 3 Taxes and Subsidies 36
Monopoly 36

Page No.
SECTION 5
COMPARISON OF BENEFITS AND COSTS IN DIFFERENT
PERIODS OF TIME 37

SECTION 6
ALTERNATIVE BENEFIT-COST CRITERIA 42

SECTION 7
RISK AND UNCERTAINTY 45
7.1 Introduction 45
7.2 Cut off Period 45
7.3 Discount Rate Adjustments 46
7.4 Expected Values 47
7.5 Game,Theory: Maximin Strategy 49
7.6 Game-Theory: Minimax-Regret 50
7.7 Sensitivity Analysis 51

SECTION 8
ALLOWANCE FOR INTANGIBLE EFFECTS 52

SECTION 9
DISTRIBUTION OF BENEFITS AND COSTS 54

SECTION 10
BENEFIT-COST ANALYSIS AND LAKEFRONT DEVELOPMENT:
AN ILLUSTRATION 57
10.1 Introduction 57
10.2 Program Objectives and Accounting Stance 57
10.3 Taxonomy of (Efficiency)Benefits and Costs 58
10.3.1Taxonomy ofBenefits from Chicago's
Lakefront Development 59
10.3.2 Taxonomy of Social Costs from Chicago's
Lakefront Development 61
10.4 Benefit - Cost Measurements 62
10.4.1.Measurement of Social Benefits 63
10.4.1.1, Reduced Flood,and Erosion Damage 63
10.4.1.2 Reductions.in Protective.
Expenditures 64
10.4.1.3 Savings in Insurance Administration
costs 64
10.4.1.4 Benefits Related to Risk-Aversion 64
io.4.1.5 Wildlife Habitats 64
10.4.1.6 Improved Recreational Opportunities 64
10.4.1.7 Income and Employment Effects 66
10.4.2 Measurements of Social,Costs 67
10.4.2.1 Construction Costs 67
10*4.2.2 Operating and Maintenance Costs 67
10.4.2.3 Environmental or Ecological Costs 68
10.4.2.4 Opportunity Cost of Lakefront Land 69
10.5 Display of Summary Benefit,--Cost Measures 69
10.6 Intangible Effect 70,
10.7 Distribution of Benefits and Costs 72

SECTION 11,
SELECTED BIBLIOGRAPHY 73

LIST OF TABLES

Table No. Page No.

5-1 Present Value of $1,000 for Various Time@
Periods and Discount Rates 39
.9-1 Distribution of Net Benefits by Income
Age, Race, and Residence of Beneficiary 56
10-1 Alternative Benefit - Cost Rating for
Chicago's Lakefront Development Plan 71
10-2 Estimate.of the New Measured Cost Per
Beneficiary Provided by Lakefront
Development, 71

LIST OF FIGURES

Figure No. Page No.

1 Demand Curve 0
2 Value of Time 33

SECTION 1

INTRODUCTION

The purpose.of this handbook is to provide concise, step by step proce-
dures for undertaking a benefit-cost analysis. It is hoped, that by
reading it the7personnel of the.City of Chicago will..better understand
how to read, interpret, and-evaluate a benefit-cost analysis; and
how to actually do one. Benefit-cost analysis is an evaluative process
derived directly.from advanced theoretical welfare economics, and can-
not be explained in detail in this report. The fundamentals-of benefit-
cost analysis.are quite'easy to grasp (although they may at first seem
counterintuitive) and if applied.with care, can provide extremely.use-
ful information on the-efficiency aspects of projects to decision makers.
Using the.procedures,:.describid [email protected] in@a report that is rigor-
ous and complete'enough for a.-Ph.D. dissertation in economics, but it
will provide numerical:results that are of a correct order of-magnitudo-.

While it-iacertainly possible,for a willing lea,rner-to be able to
assimilate.an&apply the fundamentals of:.:benefit-6-coat.analysis.in-a
relatively,short time,:it'would be naiv-eto think:he..would beable to
perform all.kinds,and all.aspects of benefit-cost analysis without.
assistance- It*would.,also,be.naive of the Cityof Chicago.to think
that a relatively short handbook will instantly-make their staff ex-
pertsAn the-1-leld. ThederiVation of deman&and supply curves, under-
taking certain surveys:and.other such activities.are best handled by
individuals.who have had specific training in:those-fields.. Concerning
such problem the goal of this,handbook will be to alert the reader
to the.possibleneed for professional help,while familiarizinghim
with the basics of the analysis so he can easily communicate with'the
professional. In some cases this familiarity.will enable him to locate
related studies that may preclude the need for professionals.

In order to get a better overview of the subject, the following is a
hypothetical example. The City is considering-redeveloping a:certai.n
tract of lakefront property into a recreational complex with a marina,
public beaches',-piznic grounds, indoor swimming pool, and several play
fields suitable for baseball, football, soccer, and other sports.

To complete themarina, it will be necessary to perform dredge.and fill.
operations, the results of which will be increased.'flood protection at
the site and other nearby areas, as.well assome ecological damage.
Current use of the land,.includes several retail establishments,.a few
private homes, some privately owned vacant lots, and a few parcels of
land that the City a.1ready'owns. The people who will ultimately have
to make-the final.decisio.n on whether ornot to carry out this redevelop-
ment will have to consider the pros and cons from many possible angles.
They will have to know if sufficient funds can be allocated to the

project. They may be concerned with the political ramifications of doing
the redevelopment at this spot rather than some other, or not at all.
They may be concerned with the-physical effects the redevelopment will
have on the environment and perhaps the. political and social ramifica-
tions that will result. They may be concerned with such income re-
distributional effects as what.will happen to the profit potential of
private firms that deal in competing or complementary products in the
locale of the proposed park, or offthe@income range of the individuals
that will use the park. They may also be concerned with whether the
expenditures for.the park are well.spent.

All of these things can be very Important and all are, to one degree or
another-, used'as background information for makinga. final decision.
Benefit-cost analysis is designed to provide information-on the pure
economic efficiency aspects of a project in terms of the value of goods
and services given up in order to build it, and the gain as a result
of its operation. It is obvious then,. that if other types of informa-
tion will be necessary or useful to decisionmakers, other types of
analysis besides-benefit-cost analysis will be.necessary.. If these
points-are kept.in,mind they will aid in the understanding and applica-
tion of'what is to follow.. The first point is worth stressing because
of the misuse of. the term benefit-cost analysis. The way the-term is
thrown around,- novices contemplating benefit-cost analysis-think:that
-all.they have to do is to list everything that looks good.and call it
a benefit and then list every thingthat,looks.-ba&and,ca11 it a:cost'.@
However, in order to provide information.on pure economic efficiency,
economists havecompiled some very specific.rules as to-what can and
cannot be included as.benefLts or costs,.and exactly how to go about
measuring them. This handbook will describe these rules in some detail.
In doing so, it:is hoped.that the incorrect interpretation of benefit-
cost analysis will not be used. To re-emphasize the second point,
just because benefit-cost analysis and, hence this report, concentrate
only on the economic efficiency aspects of a project, does not mean-that
other aspects are not important from a decision making point of.view.
Information on these aspects is much more useful when separated from
the efficiency information.

Economics is the study of the allocation of resources so as to maximize
welfare. More specifically it is,the study of the allocation of a na-
tion's resources.so as to maximize the value of the goods and services
produced.

Value is measured using the common denominator of dollars and assignments
of value are based on what people are willing to pay. For example, if
people are willing to pay $5,000 for an automobile and $600 for a-color
television,and demonstrate-this willingness by purchasing them at these
prices, then the production of three cars and six televisions has a value
of $18,600. And more important to the point of this report, if it is
possible to reallocate resources.such that the production of one car is
foregone but nine more televisions can be made, there has been a net in-
crease in the value of goods and services produced. Five thousand
dollars worth will be lost by not producing the car, but $5,400 worth
will be gained by making the nine televisions.

This is basically what benefit-cost analysis is all about; the study
of reallocations of resources to see if there,has been an increase in
the value of goods and services produced. Of course benefit-cost analy-
sis is not this easy. Often it is difficult to determine precisely
what types and amounts of goods and services are being gained and lost,
and what people would be willing to pay for them. This is certainly.
the case when evaluating alternative lakefront designs.

Therefore,a benefit-cost analysis of a specific lakefront design would
attempt to determine if the value of the goods and services produced
is greater than those given up in order to build it. Benefit-cost analy-
sis is a method for studying reallocations of a nation's (or region's)
resources to make sure only those reallocations that increase the value
of goods and services produced are undertaken. It is a decision making
tool, not to be confused with financial reports, or profit and loss
statements prepared by a private firm. As the following discussion
demonstrates, there are many differences between them as far as both
including and excluding items because benefit-cost analysis takes a
much wider social view of things than a private firm.

Regarding the hypothetical redevelopment project discussed previously,
the benefits of the project would be the value of the services provided
by its many components. That is, in order to measure the benefits it
would be necessary to determine the numbers of boaters, beach users,
picnickers, etc. who would use the project each period and what is the
value provided by doing so. The costs would be the value of goods and
services foregone by using the land, labor, and ,capital in the'land re-
development rather than in its next best use. How this is done will be
explained but note that when this information is derived it will not
provide all of the information that may be required by our hypotheticaI
decision makers. It does not show if sufficient funds for the project
are in the budget; that is a political or legislative matter. Benefit-_
cost analysis of a given project only suggests wheather or not the pro-
ject is an efficient use of funds. Another point worth mentioning is
that the cost figure derived in a benefit-cost study may vary greatly
from the actual financial cost that must be budgeted. This is because
certain items, e.g., using unemployed, labor, while involving a financial

cost, do not involve a cost in termst of foregone production. Benefit-
cost analysis will not report on how-physical effects on the environment
affect the value of servi ces produced. That is, if as a result of
dredging for the,marina, &f&ertain fish.population is diminished, then
one of the costs of the project will be the net reduction in the value
-from any c:ommercial fishery involved and any decrease-in,the value-of
the recreational services provided by the stock.

It isimportant to make three things perfectly clear at this point. Al-_
though there-may be many criteria for determining if a certain lake-
-front project'should be adopted,,(i.e., income.redistribution, strongly
felt needs of1eaders and.other@indlviduals orgroups in the community,
etc.) benefit-cost analigis.uses the criteriaof economic efficiency.
It should not.be judged.incomplete or wrong-because-of-this; it has a
Certain goal and'. -if properly used, it can fulfill that goal. If other
..criteria are-important to decision makers, then other types-of studies
will have to be done. Efficiencyis.an important criteria for govern-
ment decision makers to.use in.judging the adequacy-of a project. Even
if other criteria are used, the derision makers should know-what they
are giving up in,termwof lost-.efficiency.

Secondly, using willingness.:to-pay as.theomeasure of value is sometimes
faulted because of its dependence.on distribution ofincome. Willing-
nes-s to pay is essentially consumer needs backed-up-with dollars to
make the necessary purchases. A change in the distribution of income
will likely mean that a.different set of needs will have sufficient
backing to be registered as willingness to pay, and so the values of
different goods and services may change. Economists.have chosen to
accept the existing distribution of income as the basis for willingness
to pay on the grounds that (1) they have no criteria for selecting any
other, and (2) the existence of redistribution policies such as welfare,
graduated income taxes, etc., indicates that the current distribution
is somewhat.socially determined.

The third point is-that although.a project that increases the value of
goods and services.produced* and hence would.be approved by a benefit-
cost analysis, may in fact make certain individuals worse off. An
increase in the value of production only provides for the potential for
increasing the-welfare of society. By redistributing the gains, at
least one person.canbe made better off without.making anyone-else
worse off,,:which seems'a useful way. of measuring such things, and is
,An fact the way used in:formal economic analysis. Because a project
provides this potential does not mean that it will be achieved, For
example, using a beach area to build a marina may be an efficient pro-
jdct because the value of its,;;services are greater than those formerly,
provided to bathers at the beach, but the individuals who used the
beach will be worse off*.unless there is some way to transfer.some of
the gains to them. In many cases, of course, such transfers are not

OCLC: 4889069 Rec stat: c
Entered: 19790424 Replaced: 19950226 Used: 19930420
$ Type: a Bib LvL: m Source: d Lang: eng
Repr: Enc Lvl: I Conf pub: 0 Ctry: iLu
Indx: 0 Mo'd rec: Govt pub: I Cont: b
Desc: i Int LvL: Festschr: 0 ILLUS:
F/B: 0 Dat tp: s Dates: 1977,' %
$ 1 040 IAH Ic IAH %
$ 2 092 711.4 lb A547 %
$ 3 090 lb %
S 4 049 NOaM %
$ 5 100 1 Anderson, Lee G. %
$ 6 245 10 Chicago Lakefront demonstration project : lb benefit-cost
analysis / Ic prepared by Lee G. Anderson, Russell F. Settle, Roy F. Weston,
inc. %
$ 7 260 Chicago : lb City of Chicago, Dept. of Development and Planning,
Ic [19773. %
$ 8 300 iv, 77 p. ; Ic 28 cm. %
$ 9 504 Bibliography: p. 73-77. X
$ 10 650 0 Waterfronts Iz Illinois Iz Chicago Ix Planning. %
$ 11 650 0 Cost effectiveness. %
$ 12 700 1 Settle, Russell F., le joint author. %
$ 13 710 2 Roy F. Weston, inc. %
$ 14 710 1 Chicago (ILL.). lb Dept. of Development and Planning %

possible for practical as well as institutional reasons. It is useful,
therefore, to provide distributional information 'along with the.effi-
ciency information in@the written report of a benefit-cost analysis. This
point will be covered in some detail later.
Section 2 will provide a set of rules that if followed will result in a
useful and theoretically correct benefit-cost-analysis. It will also
contain a rather brief explanation of.the..whys and hows,of each rule.
The -following.sections will provide more detailed descriptions of some
of the important rules and will include practical methodological tips.
Section 3 contains an example of using the rules on a hypothetical lake-
front development project.

At this point, a word.is in order about,how to best utilize this-hand-
book. -In one sense, perhaps.the brief introduction into the background
of benefit-cost analysis.,andiwhat follows in the next.section are the
most important part. The former provides..the basis and underlying
-foundations for doing,benefit-cost analysis, and the latter presents a
step-by-step overviewof performing one.. Together they-provide a frame
of referenda to put the analyst.on the.proper track-for doing a study.
A common-arror,in performing-benefit-,cost-an'alyses is the inclusion of
incorrect things and.the exclusion of others. Often these mistakes are
much more important than imper.fect,.measurements of things. that actually
[email protected]. If a.person will follow the steps as set.out bearing
in mind the basic notion of pure economic efficiency, these types of
errors can.1be avoided.

The last sectio n showing.how to prepare a report with an example re-
lated to Lakefront development, is very important. The material in
the other sections describes how to make the actual measurements, and
how to handle such things as risk and uncertainty. It will not be
necessary to use every one of the tools discussed in every benefit-
cost analysis. Therefore, it:i.4 not necessary to memorize every aspect
[email protected] methodologies. It is sufficient to become familiar with their
strengths and weaknesses, the data.that they require, the ease with
which they areapplied, and the degree to which their Iresults can be
generalized. One can then pick and choose the methodologies and the
procedures that apply, either:doing them himself or obtaining the help
of an-experienced researcher, if necessary.
The goal of this handbook is toprovide theJundamentals of performing
or reviewing benefit-cost analyses. This can be defined as knbwing the
,basics, knowing how to outline a study, how to actually do-the measure-
ments.in fundamental.areas, and the ability to know when additional
help.is necessary in order to obtain meaningful results..

SECTION 2

BASIC STEPS FOR UNDERTAKING A BENEFIT-COST ANALYSIS

2.1 BASIC STEPS OUTLINE

The basicateps for performing a benefit-cost analysis are as follows.

1. Define the project.

2., List the benefits of the project.

3. List the costs of the project.

4. Determine which of the benefits and costs can be measured-
theoretically.

5. Determine which-of the benefits and costs can actually be---..
measured.,given time, budget,.and data constraints.

6. Determine-exactly what will be measured and how.

7. Demonstrate how,the things not measured could possibly-affect.
- the results.

8. Select the proper way to display the information derived so
that it will be most useful to decision makers.

When planning a benefit-cost analysis, the analyst should make sure to
use each of these steps in order to insure a complete and meaningful
analysis. When reviewing a study that a consultant has done for the
City, this checklist can be very useful as an.evaluation guide. If
the.results,do not-show that these steps were followed, the City will
not be getting as useful a report as it might have.
The remainder of this section will explain in some de@ail the whys and
hows of each of these steps.

2.2 STEP 1--DEFINE THE PROJECT

This step will specify the exact nature of the project including its
purpose, objectives, and its relation to existing and,potentially re-
lated projects, and the 'appropriate accounting stance to use while
undertaking the analysis. It is necessary to ensure that, the,study will
focus on those things most important to decision makers, but a benefit-
cost analysis can be very expensive and it is important to frame the
study so that the proper information can be obtained as efficiently as
possible. 'Determining exactly what the decision maker is likely to need
is a necessary part of an efficient analysis.

The first two parts of this step, specifying.the purpose and the objec-
tives, are included to explicitly state the what and thewhy of the
project. The what shouldinclude a detailed description of.the project
or the program to be undertaken. The why should include a statement of
thereason,for doing it.

In most cases, specification,of the purposes is quite elementary,, being
spelled.outlin:the charge@to theanalyst. With multipurpose projects,
however,- only one or two,may be listed, but it i&important that all be
studied.. This will insure that the decision-maker is apprised of all
of the consequences,of the7,project. Determining the objective should'
pose little difficulty. The-purpose of benefit-cost analysis.is to
determine the efficiency-aspects'of various projects; therefore,.-for
purposes of.the study..the.main objective,should be-to increase the
value-of the:goods, and.services produced. In some cases decision-makers
may have,other objectives such@aaincome_redistribution, or. regional
development... These.other objectives-are most frequently used in de-
veloping countries, but.,the City,of Chicago@may consider them in lake-
-fronvdevelopment planning. In this case, the analyst should structure
his work.so-that the report wUlindicate to what degree these objectives,
have been met.

if other.similar projects or programs are,being undertaken or.contem-
plated by,other levels@of government orin thepr ivate sector, an exact
statement of:the relationship between them should be included,:in this
introduction. For example,Af the state government is planning to open
a new park with easy -access1for city dwellers, or if a well-kept vacant
lot that is often.used as a playground is going to be-used for a con-
struction site, they would have important ramifications on the useful-
ness of a planned lakefront park developed by the City.

The accounting stance,is the frame of reference for looking at the pro-
ject to be studied- Benefits have, been defined as people's willingness
to pay for the outputs of the project; and costs as the value, as
measured by.willingness to pay, of the foregone opportunities of doing
the project. Before-implementing these definitions, it must be
specifie&exactly who.will.be included in the calculations. That is,
should the willingness to pay .be included only for the residents of the
City of Chicago or expande&to county, state, regional, or national
levels.

If a completely general accounting stance is adopted, the rule is to
count.aIll benefits and costs to whomever they-accrue, where benefits
and.costs are as defined above.-. As the accounting.stance is narrowed,
the.rule,is to count only those benefits and costs that accrue to,
individuals specified in. the frame of reference. It should be obvious
that the accounting stance should be stated very explicitly, so that it
is easy to.determine.just whose benefits And costs are to be counted..

Selection of the accounting stance is quite important, both in deter-
mining the-cost of doing the project and in the type.of decision that
is ultimately reached. If the City of Chicago takes the position that
it will count benefits and costs that accrue to or are borne only by
its citizens, this will confine the analysis and make.it easier and less
expensive to-perform. From a narrow maximization point of,@@View. the
City may be quite wise in considering only the gains and losses to its
citizens when making decisions on:how to spend its tax revenue and its
land and water resources. There is basis for the argument that unless
the.benefits to-the people who are paying for the.project are greater
than the cost, the project should not be approved. Nonetheless, because
the effect of lake front projects can be felt state and even areawide,
a.narrowly defined accounting stance ma .y lead to inefficiency as far as
the nation is concerned. For instance, in the process of using landfill
to obtain new beach area, the lake may beecologically affected in such
a way that the value of services it provides is-decreased. If decrease
is primarily to nonresidents, then a narrow@accountlng stance may approve
a project actually leading to a net decrease in the value of goods and
services produced-.as far as the whole area is concerned. While the
ana:lyst cannot always determine-which:accounting stance to use,.-he must-
make it known if the onlyreason a c 'ertain project is approved-is-be-
cause a narrowview has:been-taken.

For lakefront.development a narrow accounting stance should be sufficient,
especially-withregard to measuring benefits. The great majority of
the benefits of a local park will be to city residents. Using such a
stance will save money, make the workload more manageable, and yet should
not significantly alter the basic results. This approach should be used
with caution, however.. In those instances where significant benefits
or costs occur to nonresidents, a wider frame of reference will be
necessary to get a true picture of the efficiency of the project. Even
so the ultimate decision maker may decide to look only at the local bene-
fits and costs. This is a political decision that may result in bad
economics from a national point of view, but as long.as the decision
maker is aware of these ramifications, the analyst should use the
accounting stance that is requested.

Although project definition is a very important step in undertaking a
benefit-cost analysis, it is one that the novice can perform with little
difficulty.

2.3 STEP 2-LIST PROJECT BENEFITS-

A project may have many effects, butfor a strictly defined benefit-
cost analysis only those that will affect economic efficiency need be
considered. A good rule is to define benefits as the amount that people
are willing to pay for the.direct outputs of the project. The,..,@key
phrases are: -"willingness.to pay" and,"direct output." According to
economic theory, willingness to pay is the proper measure of value to
use when ascertaining the worth of something (see Section 1).

Direct outputs are those things actually produced in the construction
and operation of.the project. It is at-the same time narrower in one
waythan, one might expect,but..broader,in another. It is broader be-
cause it includes,all outputa, not just those for which the project was
planne& . For example, a.marina outputs dockage and storage,facilities
for boats, potential environmental damage as a result of dredging and-
boat operations, reduced-visual enjoyment of the.lakefront, etc. Obvi-
ously,, the marina was built to,provide the boat storage and, dockage ser-
vices, but as a-direct result of.the other outputs, which are negative,
benefits occur. To be complete, they. must be counted.in a benefit-cost
analysis. In many casesi these outputs can be larger than the ones for
which the project was designed meaning that ignoring them will lead to
approval of a project.that.act@ally provides net efficiency losses. It
is quite difficult, sometimes almost impossible, to get a perfect'meas-
ure of willingness to@payfor,these noroarket outputs, but as-the next
.discussion will show.- methodologies to do so.exist.. In spite of these.
the nonmarke't outputs.musi@be-included in a benefit-cost.
analysis.. Their-inclusion1helps differentiate benefit-cost analysis-
from normal business.profitability analysis.

The definition of benefits as willingness to pay-for the direct outputs
of the project is narrower.because it does not count:such things as the
extra services provided by motels that locate next to the marina, nor
the revenue earned:by a-hot dog salLismaimworking the area, nor-the.ef-
fects of the spending.habits of those who construct or operate the.fa-
cility. These things, called secondary or parochial effects, are not
really net benefits in the sense that the value of goods and services
is being.increased. If the economy,is.fully employed, the resources and
labor to provide these items are p .ulled:out of other areas.where.they
could provide-services,of equal value. If one is.interested-in the eco-
.nomic impact of-the project,,these things should-be considered., but for
all practical purposes, they are notrelevant for a benefit-cost analy-
sis. The phrase "for.all practical purposes" is used because there:are:
certain situations where these items.will.be important.' The-most im-
portant case.is the use of unemployed resources. Because they are, by
definition, producing nothing, their use does notinvolve a reduction
in goods and,servides produced;.they have a.zero opportunity cost as far
as benefit-cost analysis is concerned... This concept will be discussed
in some detail later. In most cases, the best rule is to ignore them.
This is especially true for projects,the size of those envisioned for
lakefront redevelopment.'and in situations where the.potential analyst
has only arudimentaryunderstanding of advanced economic theory. The'r6-
fore, it is recomme 'nded-that when analyzing potential shorefront designs,
benefits'be defined as,the total amount that people.are willing to pay
for the-direct,outputs,-of the project. All other effects.should be ig-
nored. It -should.be emphasized, however, that when narrow accounting
stances are used, direct outputs.may include more things.

For exampleAf a local accounting stance is used',.an increase in related
business activity in the Chicago area can-be considered a direct output.

If a national accounting stance is used,.this would.not be the case be-
cause with full employment-an increase in business activity in the
Chicago area woul&be possible only with a decrease elsewhere. From
this perspective *there would.be no direct-output at all as far as related
business activity is concerned, only a change in its relative-16C.'ation
(assuming'full employment elsewhere).

This step in preparing a taxonomy of the project benefits is an expan-
sion of the first part of step L. The purpose is to be very explicit
about what will@be produced-as a result of the project. At this stage
the analyst should not be concerned with the expected.size of any of
the benefits.or of the ease-or difficulty with which they can be meas-
ured. Every possibl:e direct output benefit should be listed. This
provides a good overview of@ the benefits of the project and@provides a
checklist-for the analyst mmeasure his success in completing a.full
analysis. A novice benefit cost analyst should have no trouble perform-
ing this stepi It is only necessary to keep the definition-concerning
direct outputs in mind.

2.4 STEP 3 LIST PROJECT COSTS

This is the counterpart of step 2. The concept of opportunity cost.is-
most important. What we are looking for is what things aregiven up in
order to. undertake the profect, and what is the willingness to pay for
them. That is, costs are the foregone benefits from not being able to
put the resources to their next best use. To be complete, the last of
costs should include an evaluation of the resources necessary to plan,
.Construct, and maintain the project as well as any to ensure compliance
with its use. The analyst must decide the types and sizes of resource
reallocations by both public agencies and private individuals and fix
as a direct result of the project. The next section and the hypotheti-
cal example will clarify this step.

What is called for.is a complete list of the amounts and qualities of
the actual inputs (i.e., land, labor, and capital equipment) that will
.be used in the construction and operation. Also costs.of the project,
in the sense that some good or a,servicewill no longer be produced
(even.if they will not be a part of the financial cost of building or
operating theproject), should be included as well. In some instances,
it will be more appropriate to include the negative benefits described
as a cost just as long as they are notlincluded in both 'lists. Like
step 2, the novice*analyst willbe able to handle this step.

2.5 STEP 4 -MTERMINE WHICH BENEFITS AND COSTS
CAN BE-MEASURED THEORETICALLY

From the definition of a benefit,it should be clear that while it may
be possible to conceptualize'a proper me .asure, actual application may
be quite difficult., Measures of willingness to pay for items that are
freely purchased and sold in themarketplace are quite simple to obtain.

Market price paid is a useful proxy for this if an item is not worth
the pride, people will not purchase it. In some instances, however,
adjustments are necessary. For example, if the project will expand the
output of a product by a large amount.relative to that already sold it
will in all probability, have an effect on the price. in order to
measure benefits.proper ly.it will be necessary to know the demand curve
for the product.in.order to predict the exact nature of the change . @in
price.- It is also important to [email protected] a:price change occurs,
an extra benefit is provided-to existing,consumers because they can pur-
chase the same.product at a lower price* Another case where themarket
may not provide-useful.,measured of benefits or costs is when there is
monopoly-or monopsony in-the market for inputs or final outputs. To be
specific, the real cost of using-an input produced by a person with some
monopoly power is-less than the price he charges. An extra amount of
revenue,in that.price is due to-his market power. The measure of cost
needed-is the value of,.goods,and services that could be produced if the
monopolist ceased product ion@and released-the:resources for other uses
in the economy.
The problem of measurement is.much more difficult for the goods and ser-
vices produced or.given up that-are.not.traded on markets. For example,
the operation of a.power plant may.produce negative benefits.by lowering
the air and water quality in an.,area which may reduce,the amount and
value of-services provided by,the environment. The lost services may
include everything from logt-fisheries.productivity'to decreased enjoy-
ment from evening..walks. Onthe other hand, the benefits from past
scientific investigations into polio control were reduced mortality and
morbidity,

The purpose of this step is to take the lists developed in steps 2 and
,3, break thezinto the market and normarket categories, and determine
which can be measured given existing theory. Some of the important
methods used to measure thetypes of benefits and costs likely to be
associated with lakefront development will be discussed in the follow-
ing sections.

'If it appears that.the majority of the-benefits orthe costs are.of the
type that cannot be,measured-using.existing social science research
methodology,. several.preliminary conclusions@may be determined at this
point. First, unless one of the measureable benefits is very,high and
the project appears to be a clear winner even in,the presence of all of
the nonmeasurable effects, the benefit-cost analysis may not result in
a definitive Answer. This may mean that the best one can do is to pre-
sent the results of, the steps so far, hoping that they may be of some
use to the decision-mdker. Alternatively,lt may be a signal, that in
order to,significantly improve the report it.will be necessary to brin .g
in a professional consultant.

Familiarity with the,next two sections will allow the novice to make a
very good start at.6
ompleting this step. No@actual measurement it re-
quired, only,a notion of whether the.benefits.and costs are amenable to

measurement. Since the discussion of ways to measure benefits and costs
cannot be complete (as far as including all types of measurement or how
to apply each of these in every possible way), there may be instances
where the novice using this handbook may make an-incorrect c1assif -ication.
The effects of this can be minimized by classifying a benefit as poten-
tially measureable if there is some doubt about its relevance-to existing
methods. If nothing can be found to work later on in the study, then
this particular benefit can.be reclassified as nonmeasureable.

2.6 STEP 5 DETEMINE WHICH BENEFITS AND
ZOSTS CAN BE MEASURED

Probably if.enough manpower, computer facilities and other related re-
sources were used to evaluate all the benefits and costs of a project
,-(market or nonmarket), a suitable and acceptable-measure could be ob-
tained. The problem is., that the cost of doing.the analysis if very
-important. In most cases-there are critical time, budget and data con-
straints. The analyst must plan to provide@the most useful information
-given these constraints. This process involves much more@than merely'
measuring all the things that can be.measured-until the-cohstraints are
met. -It.should be a process of determining-which benefits and costs are.
likely.-to be significant and.then trying to obtain as precise a measure-
ment of them as possible to determine whether benefits are greater or
less than costs.

In some instances the items that are difficult to measure may be so
minor that it is of little value to spend time measuring them. In many I
cases, however, they will be central to the analysis and can be ignored
only at the peril of*making a wrong decision. Other times it is diffi-
cult to determine.in advance whether the difficulty of measuring items
will have a significant impact. It is here that the integrity of the
analyst is critical. To avoid errors it is advised that he should adopt
the policy of attempting to make measurements if there is some question
about the relative impact.

Essentially, step 5 is a further breakdown of one of the categories de-_
rived-in step 4. Of all the effects, only a number may be studied.in
detail because of various constraints on the analyst. The problem is to
make sure that the effort spent on measurement is profitable.

2.7 STEP 6 - DETERMINE WHAT WILL BE
MEASURED AND HOW

This step is the logical conclusion of steps 4 and 5. The criteria for
determining what should be measured are:

0 Awareness of relative importance of the i tems.
0 Possibility of obtaining comparison values from other studies.
19 Existence and difficulty of applying measurement tools, etc.

As an interrelated decision, selection of the measurement procedure
should be based bn the exact'nature of the items to be measured, and
the budget and data constraints.

,Depending.,on.the nature-of the project to be studied and the speed and
degree of accuracy requitedi-a.,novice analyst may be able to complete
this step as well.- If most of the effects are-fairly straightforward
and there are,,fewexternal'or.-other nonmarket effects, then the method.!-
ologies.discussed in the,following sections will provide an ample basis
for performing,a very credible.report. In some cases, even though there
-are,complexeffects to be measured, the problem of actually making-meas-
urements can-be bypassed by drawing.on.values,found in economics liter-
ature or in other credible benefit-cost analyses.@ In both cases it is
important.that the.analyst review the work to make sure that it is based
on proper.economic methodology focused on, the appropriate variables. In
this@lregard,, this handbook will,.provid.e a useful frame of.reference. for
making such decisions.. Also it is crucial to make.sure.that the situ-
ations being-compared are.similar enough solthat meaningful adaptations
can be made. For-example,. a study.on the value of recreational services
-in-Wyoming-will be-of.little-use in placing.a unit value on@a recreation
day on the-Chicago lakefront,

In some other instances the,novice must realize that there-are special
problems that.will require professional consulting help. This does not
mean that@the novice must completely.give:up at:this point. With his
knowledge of the problem and of benefit-cost analysis, he will be able
to tell exactly where professional,help'is.needed, and will be.able to
get the consultant started immediately by making the work already done
available to him. It is important that the novice only use the consult-
arit for those things.which he is not trained to do. For example, the
consultant should*be requested to estimate-the demand curve for a par,
ticular output, but the analyst should.be able to utilize the curve to
perform,the rest of the analysis as described in this handbook.

2.8 STtP 7 - DEMONSTRATE,EFFECT OF,
NONMEASURED ITMIS

In some instances, the things that cannot be measured for theoretical or
practicable reasons can affect the results of the study. If so,.the
analyst must at least specify that certain things have not been measured
and-.indicate whether it is a benefit or a cost and,.if possible., the
order of magnitude involved., For example, excluding a measure of the
recreation servicesforegone by,using an area of beach for a marina fa-
cility underestimates the costs, and, will im prove the chances of the
project.being accepted. Thisis especially.important to the decision-
maker if the difference between measured-benefits and costs is,quite
small. With a known cost not being measured, he will then have some
reservations.about making a positive decision concerning the project.
On the.otherhand, if the benefits far outweigh the costs, then the
decision-maker will not be so worried about the unmeasured costs.

Obviously, if only a few effects cannot be measured, then this type of
analysis will be quite useful, because it will give some idea as to
whether the overall evaluation is overly optimistic or pessimistic;
i.e., if benefits are overestimated and/or costs are underestimated or
the reverse due;to the existence of nonmeasureables. If there are many
things that cannot be measured, and.their cummulative effects could be
in either direction, not very much information hasbeen added. It is
still-a good idea to perform this step so that-the general tendencies
may be made'known'to -the decision-maker. This step can adequately be
performed by a novice.

2_9 STEP 8 '-. SELECT- PROPER METHOD OF
DISPLAYING INFORMATION

The final step of a benefit-cost study is actually writing the report.
The presentation of the benefit-cost findings is very important since it
serves as theprimary channel of communication between the analyst and .
-the decision-makers. The report must be carefully written and the find-
ings properly-Aisplayed, or valuable information may be overlooked or
misunderstood.

The first step is to clearly identify'the presumed objectives-of the
program or.project being analyzed and relate those objectives to the
stance adopted in the study. It is-especially important that
the.study's viewpoint, or accounting stance (e.g., citywide, regional),
-,be clearly identified as this information is valuable to policymakers
in judging the relevanceof the study's findings.

Next is identifying every benefit and cost that should be considered and
where possible, measured. This taxonomy provides a benchmark against
which the analyst's success in quantifying a project's benefits and costs
can be assessed. Even though it is clear from a study's outset that cer-
tain effects are unmeasureable, they should be included in the benefit-
cost taxonomy.

Once the taxonomy of benefits and costs has been developed, the approach-
es used in measuring those benefits and costs that were susceptible to
quantification should be discussed. This section of the report should
emphasize the more restrictive-and controversialassumptions made in
constructing the various benefit-cost measures.

The next recommended step isito display the summary benefit-cost statis-
tics, either net present values (the difference between discounted bdne-
fits and discounted costs) or benefit-cost ratios., The information
displays should convey concise information about a project's estimated
effects (benefits and costs) and present the results of various sensi-
tivity analyses to aid decision-makers in identifying the study's crucial
assumptions.

Finally, the presentation of the measured benefits and costs should make
clear that there may remain important unmeasured effects and that the

distribution of the project's benefits and costs may have important
policy implications.' 'Gridd the summary benefit-cost measures are pre-
sented, they must be carefully qualified.to the extent that there are
potentially important intangible or distributional effects associated
with the project. These_issues will be discussed in more detail. The
novice-can do this quite adequately,, even if it is necessary to hire a
consultant.,for one of the particular measurement tasks.

Essentially a concise and well structured report should.-provide a fairly-
detailed description of how the first sevenateps of performing a
benefit-cost analysis were accomplished. The last section of the hand-
book contains an illustration of aproper benefit-costxeport.

SECTION 3

PROBLEMS.OF MEASURING BENEFITS

3.1 OVERVIEW

This section will provide a fuller appreciation of thedefinition of
benefits and introduce some@of the more common methodologies-for ob@-
taining proper measures. Benefits are the total willingness to pay
for-all goods and services that are direct outputs of.projects. Will!---
ingness to pay is accepted as the proper way.of measuring value because
it is the common yardstick used in market economies. To ensure that
problems, such as environmental pollution are included (which cannot
be adequately handled in an unregulated ma ket economy.) and yet-to in-
sure that double counting.and inclusion of offsetting values are omitted,
all-of the directoutputs, but only direct outputs,'are to be counted'..
Ifan analyst can keep.these,two points constantly in.mind as-he selects
tasks and properly measures the correct benefits, a proper benef-it-cost
should result. These points should be the basis for evaluating.
---other--benefit-cost reports, because-it will be possible to dismiss-those-
-that err in measuring the wrong things or use an incorrect measuring
method. 01
Listing the direct outputs@and then determining which theoretically and
practically can be measured are the main steps in a benefit-cost analy-
sis. The last part comes down to [email protected] demand curve, or at
least a portion of a demand curve, can be established for the output
in question. The demand curve is.an amazingly useful theoretical con-
@cept as well as a very useful tool. It is at once more complex than
the uninitiated might suspect and yet easier to apply than might appear.
The reader should not become disillusioned at the abstract nature of
next discussion and wonder about its place in this handbook. In many
cases it will not be possible to obtain precise estimates of demand
curves for all of the outputs,of lakefront redevelopment programs.
Therefore, there-will be nothing in this handbook that will teach a
novice how to derive a perfect estimate of the demand curve in every
instance.

However using the admittedly difficult-to-apply concept of the demand
curve allows a firm frame-of reference for benefit-cost analysis. If
one could be estimated in every case, the problem of doing a benefit-
cost analysis might become as simple as adding,a column of numbers.
But keeping in minda complete concept of a demand curve and what it is
trying to measure will focus research efforts in the right direction.
Used properly, it@will insure that only proper effects are considered
and that @here is no double counting.

The remainder of this.section will define demand curves and the related
concept of consumers" surplus. We will discuss how demand curves are
related1to existing market prices and under what situations market prices
are appropriate as measures of willingness to'pay. There@:,Vill be many
times.when this is,not appropriate, especially for lakefront develop-
ment projects:2.and procedures will bedescribed to corr 'ect for this.
-In some cases, these descriptions@will serve as actual cookbook-type
instructions for performing the procedures., In other cases%it will
merely-be possible to describe their strengths.and weaknesses,.the type
of problem tawhich-they are suited, and-then show-how the analyst can
work with a professional if it is felt that a particular procedure is
necessary. It is.important to-remember that each procedure is nothing
more than an attempt to:derive a measure of willingness to pay.

Finally,. note that the.proj,ect.cost as far-as formal benefit-cost analy-
siais concerned is the benefits:foregone by precluding.the use of the
resources,n.ecessary toundertake the project from being allocated to
another use. The use.of this means the labor and material to produce
.-a-specified lakefrout development project preclude&their use elsewhere
in the economy, The value of what these resources@could have produced
is-the,opportunity cost oftheAevelopment. For example, if one--of--the
inputs-is:-cement, then in order to determine which people are willing
to pay for,cement for another purpose, it is necessary@tostudy the
demand curve for cement. Because opportunity c.osts--are.the,foregone
benef.i.ts,caused by precluding the use of input&elsewhere, the following-
discussion can be applied to.the measurement of costs as well.. It is
only necessary to reverse the frame.of reference and consider the demand
curve-for goods and services foregone.. Section 4 will.describe some of
the special cases when measuring costs can be more difficult.

3.2 DEMAND CURVE: MARGINAL AND NONMARGINAL CHANGES IN OUTPUT

The basis for estimating willingness to pay is the'demand-curve. A
demand curve can be defined as the relationship between the price of
a product and the number people will purchase per perio&of,time. It
shows the maximum amount that'can,be obtained on the market when various
amounts are provided in a given period.

In Figure 3-1, when 500 units are provided each period, the.market price
is sa. This means that willingness to pay for these 500 units is at
least $4,000 because.,that i's-the amount paid on the market. The total
value is somewhat greater than this because if the price were raised,
the quantity demand,would fall, but not to zero. Some people would pay
more than.the $8 market.price. :rhis extra amount, called consumersi
surplus, can be represented by the.area of the triangle ABG@2 Since the
amount actually-paid on the market equals the area of rectangle CE, the
total willingness to pay for 500 units per period is area ACEO.

.The importance of the concept of consumerst surplus can be explained
in the this way.. If a project will increase output of the hypothetical
good from 500 to.700, then market price will fall.to $6. How can will-
ifigness to pay-71br,-these extra 200 units be evaluated? Is the proper
price $8or $6.? A useful way to look at the problem is to divide the
..total willingness to pay-into what is actually paid on.the market and
the increase in consumer surplus, The 200 units sell-for $6,-each or
$1200. Note also that the area of consumers"surplus has increased by
BCGI.- Of that, BCFI represents the decrease in price to existing con-
sumers ($2 times 500 units or $11000) and CFG represents-a bonus to new
consumers in the sense that they are paying something less--than they
actually would be willing to pay. Assuming that the demand curve is a
straight line, this amount,can be obtained by multiplying 1/2.times
$2 times 2Wor $100 The total willingness to pay for the extra 200
units is therefore &i.C11"" Of these three amounts or $2300.

In &different context, assume that there is a lakefront redevelopment
project-.that will, in part, provide.200 new-berths to moor pleasure boats.
This is@a direct output*of:the,project, and so, in order to complete
the benefit cost analysis, we must find out exactly-what people are
willing to pay for it. Assume that the demand curve pictured in-Figure
3t-Lis the demand curve for berths. @urrently, there are 500 available
and so the price will be $8.. We can tell from the demand curve that.
when.the extra-200 berths are made available,.the market price will fall
.to $6. Therefore * these extra 200 units are worth at least $1200.. Note,
however, that the users of the original 500 berths now face a price of
$6 rather than $8. Therefore they-would be willing to pay at least $2
each to see the redevelopment project undertaken.. This amounts to a
willingness to pay of $1000. This is still not a complete picture be-
cause some of the people who rent the new produced.berths, at the $6
rate, are paying less than they would pay rather than do without. The
individual who would-be most eager to rent would be willing to pay
something just less than $8 while the marginal consumer is just bare "ly
willing to pay the $6 rental fee. The others are willing to pay some-
thing in between. This is demonstrated by the shape of the demand curve
between points C and G. An estimate,of the consumers? surplus for the
new consumeracan be found by averaging the extra'willingness to pay
over the 200 units. This is the same as finding the area of triangle
CFG, which is 1/2 times $2 times 200'or $100. The grand total of will-
ingness to pay for these extra 200 berths is the sum of these three or.
$2300.

This is an example of how.to measure total willingness to pay when the
output of the project is large enough relative to current production to
force a reduction4@in mark6t,price. This will not always be the case,
of course. If the existing number of berths was 50*,000 ratherthan 500,
the extra 100 would 'probably.Ihave very little, if any, effect on the
market Price. Therefore, after the completion of the project, the
price would still be $8. -The willingness to pay for the extra 200.units
would then be $1600. With no decrease in price, there would be no con-
sumerst'surplus, and so the $1600 would be the total willingness to pay.

The practical lesson to remember from-this analysis is that market
price is a measure of willingness to pay, and for those direct outputs
that are sold in markets, it can be a meaningfulmeasure if the extra
amount supplied-is,so small relative to the exigrbing amount that no
change in price is expected. In cases-like this, the work of the analyst
is quite simple.. :There is:no need to'know the entire demand curve,. but
only the existing price. When,,the extra output is large enough that a
change in price will follow,,it is necessary to take into account in-
creases in consumer surplus as well as,.any revenue earned on the-market.
In these cases it is,necessary to have informationon the demand curve,.
and if no-,estimates can@be found in the economic literature, then the
services of a qualified econometrician will'be necessary.

After discussing the nature.of demand.curves and consumerst surplus,
it is important to see some.restrictions on their-use,,and suggested
ways of estimating.willingness to,pay when a,.formal d :curve cannot
be:obt4ined.. A word.of,caution here is in order.- It@is not necessary
to use every one.of the following-procedures on every-benefit-cost
analysis. In some cases.,,. the method's will,not be@relevant because
standard-market@[email protected] appropriate, and iniothers @the:pro'blem:
may-baso small that a known.,deftciency in demand-.analysis can be safely,
ignored. For.example, if-one:tenth.of one percent of the dollar co 'sts-
of a project are producediin a market where there are price ceilings
(see the next section), it wil1not be worth the effort to find the
difference between the market price and.the correct social measure.
The following discussion should be considered a list to consult when
undertaking steps 4, 54and 6 and a source of information if any of the
methodologies are deemed relevant to,the,project.

3.3 RESTRICTIONS ON DEMAND CURVES

In cases of institutional interference in the market, the use of observed
prices is not the, @proper measure to use in benefit-cost analysis, even
when there is only a very small change in output. Price ceilings, price
floors, and output controls are themost notable examples. It.is possible
to correct for this in a relatively straightforward manner.

3.3.1 Price Ceilings

Price ceilings are government decrees that the pride of a certain product
cannot go above a specified limit. Rent control is a prime example.
A generally accepted maxim directly related to lakefront development is
.that access to lakefront recreational facilities should.:be free. In
essence the price ceiling is set at zero. A further example would Ibe
the case of a fixed price for mooring spaces belo,4-the price thatwould
exist in an unrestricted.market. Whatever the product, if a price
ceiling@is binding (i.e., if the price would,tend,to be higher without
the ceiling), the visible market price is not a reliable estimate of
willingness,to pay. Because people do not have to pay for the use of
beaches does not mean that no value is-being provided. Extra apartments

DEMAND

F G
6

E H
0 500 '1100 Q/TIME

F I G U R E1 0 E M A N 0 C U R E

built in an area with rent controls will providea value to consumers
greater than the amount they.,pay in rent. In these instances, the ana-
lyst should strive to obtain an estimate of the-true willingness to pay.
An example of,how to do this for recreation project follows.

Price ceilings also have an affect on the 'measurement of costs. When
an input-is producedin a market with a price ceiling,.then as long.as
the amou%ituaed will-be the result of new production and if the new
production will not significantly affect total market output, the price
ceiling,is an-adequatemeasure of cost. Producers-will not produce it
unless the regulated price at least covers the.cost of production. The.
price ceiling-.merely,separa'tes the amount.that people are willing to@
pay from the regulated.market-price; it does,hot-affect the amounts
people will sellat different prices. Where the inputs are taken from
current.production, then the-willingness to pay rather than the regula-
ted.market price is the proper:measure. It should-be noted. that in
most cases any inputs from a market subject to price control.used on.a
project will result in 4 decrease-to current users because suppliers..
will be producing all they-care to at the fixed price.

The-reasoning behind this somewhat.paradoxidal rule is the basic defini-
tion of opportunity.cost.. If'the:output is fromnew productiont the.
market price.is the correct measure because it measures@the cost of
J- -is, of. course, lower than willingness to pay.
extra production (wh.-h
because of the price ceiling). If thetoutput.is taken.from current
users, the willingne@-,s to par is a.measure of the value given up.

As an example, suppose that a lakefront project will use 1000 gallons
of gasoline to operate heavy earthmoving equipment and that as part of
a conservation-inflation program the government has fixed the price at
50C a gallon. At this low price the fuel companies cut back on their
production, yet consumers.want to-purchase more. Therefore, in order
to make the program work, rationing will be necessary. In the presence
of the rationing-system a black market in gasoline-develops where the
fuel sells for 750. If the City of Chicago is able to get the 1000
gallons it needs,. how should the cost be measured?- Although the budget
cost,of purchasing the gas will.only be $500, the actualsocial oppor-
tunity cost will be $750 because that is the amount other people would
be willing to pay for the gas if they ha&the, opportunity-

The main point to remember when dealing with pricaceilings is that the:
artificial price is not a real measure of--what people are willing to,
pay. It is incorrect.to use the:fixed price as a unit value of outputs
of the project or for the inputs used to undertake it;,another measure
is necessary. The existence of black markets.is,a good way to obtain a
proper estimate, as is the method for estimating recreation demand-which
follows.

3.3.2 Price Floors:

Price floors.are@the reverse of price ceilings. They are prohibitions
that prevent the price of a particular good frovr-fa@lling below--a set
minimmi. Agricultural price,supports and-minimim wage laws,are.common
examplea.- If the-price support is binding (i.e., it keeps-the price
higher than irwould otherwise be), then-the regulated price is not a
suitable measure of willingnessto-pay. @This.point can be clearly seen
-.--by using-the example of price supports in agriculture. In-order to main-
tain a price.above that which-.can.-belrear-hed-without intervention, the-
govertment must buy the."surplus" or the difference between what con-
sumers will purchase at that price and the amount farmers are willing
to produce. In this@instance, it-is easy to see that extra production
by a proposed.project surely should not be valued:at the regulated
price because-there is more..than consumers-want at that.price..-Th6
problem is to find out.what use is made-of.extra.goods produced and@._
then obtain a measure of,what people are@willing to pay for it in that
-use.. In the case-of government storago-of farm surplus,.the value is.-
negative because resources must be spent to store or destroy-it.

--To measure that cost of-using inputs produced in a marketsubject to
price floors, the source of the inputs must be determined. If-they.-_
@only-reduce the surplue, the cost is.negative because no.extra.resources-
are-used and storage costs are eliminated. If they reduce the-amount-
available to consumers or if they come from-newproduction, the cost
can be measured by the regulated market price. In the former case-the
r6gulated'market price is the correct measure because it is what the
people who do buy the goods are-willing to pay (i.e.,.it,is the amount
they would lose if they could no.longer consume it). In the latter,
the regulated price is a proper measure of cost because.it is the'amount
necessary to entice producers to expand production. (See example at
end of next section.)

3.3.3' Output Controls

Output controls are regulations prohibiting production of a certain
product beyond:a certain point.. They are quite similar to price floors
except that the limit is on output rather than price. With price floors
the price is held upby government purchase of any surplus beyond what
consumers want at that price. Output controls accomplish the same thing
by preventing the growth of output beyond the amount consumers want at
that price. Output aontrols.have no surpluses with which to deal.
While the regulated price is a correct measure of what people are willing
to pay for extra units of this good, if the output control.is binding
upon the project, it is obviousthat the benefits",will be z6ro because
extra output will not be permitted. The cost of an input produced in a
market with output controls is the market price because any units on
the project will have to be transferred from other uses as long as the
controls remain in effect. If the amount used omthe project is.large
relative to the'total market output, then losses in consumer surplus
will have to be included as well.

It is doubtful that price floors or output controls will ever be-a
problem in estimating benefits from lakefront development. They may,
however, raise difficulties in measuring costs. Assume that Q in@the
demand"etwove in Figure-3-1;represents@an input for a project. If
production is-1imited by government order to 700 units, then the price
will be $6. This:is the measure:that should be used for cost when the
project will take only a sma-1l.percent,of total production, regardless
of howmuch-it.actually costs to produce:the-item. This is the amount
that people are willing to pay f'or'each unit,,and is the measure of
whatthey will loseif the inputs-are reallocated to the project. In
the casewhere the project will.use,a large portion of the-amount
supplie&,then-the losses in@consumer surplus must be considered. For
example, assume that the project will use-200'units of Q, and in doing
so it-will reduce theamount available to other consumers to 500.
Therefore,,the total loss [email protected] area BCGHFI, which is
equivalent-to $2300 as was,explained in the discussion of consumers'
surplus. It should be obvious@that this is just-the reverse of pro-
viding 200,units of Q.because [email protected] result..of project production,@200
units of Q will.no longer:be available to.-the public. Therefore,- the
net result is a $Z300 cost ratherthan a.$2300 benefit.

3.4 OTHER MEASURES OF BENEFITS

In those cases where the direct outputs of the project are goods and'
serlvices.n6t normally sold on the market, such as.recreation or pollu-
tion causing waste products, other methods must be used to estimate
willingness to pay. This section will review many of them. A reference
supplied by the planning department entitled "A Preliminary Guide for
Benefit Considerttions.in Economic Impact Studies" has a short section
on this subject. Since it is.assumed that this work will be available
to city workers undertaking or studying benefit-cost analysis, it was
felt that it.would provide a.useful framework for this discussion.
The book discusses methods but we will discuss only those that are
directly relevant to lakefront development. A quotation.from some of
the relevant.sections will be presented, with a brief discussion of how
the method fits into our analysis, and to emphasize important points
or add new material if necessary. Another method not described in the
IIEQ study will be discussed.. It should be remembered that it is not
necessary to use each of the.following methods in every benefit-cost
analysis. Theimportance of each method depends upon the exact nature
.of the problem'under.study.

This material is.intended to help measure benefits, and methods will
havedirect application to lakefront redevelopment, but in many instances
it will be,,,,in terms.of'costs. That is, a certain project may not re-
duce environmental pollution but its.construction may lead to some.
The following quotes will have to be interpreted with this in mind,

"'A.Preliminary Guide for Benefit Considerations in Economic Impact
Studies," Governors State University, Park Forest South, Ill., IIEQ
Document No. 7.6/12, June 1976. 23

3.4.1 Monetization of Value of Materials
Deterioration Resulting from Exposure to Pollutants

Afteradjusting the economic value of the-material with respect to
economic life and the percentage exposed to pollutants, the "value of
interaction" is calculated by.eAtimating.the difference between the
rate of materials deterioration in a polluted versus unpolluted environ-
ment. The adjusted economic Value is then multiplied-.by the value of
interaction to yiel&the value of materials deterioration.

Data on th6effects of pollutants on many materials is-unfortunately
non-existent, and the sources-and types.of'pollutants.with which the'
materials are in contact at various times are in most-cases difficult
to identify.

Thi's general type of approach is directly relevant to studies of changes
in the structure and formation of the lakefront. Pollutants can often
cause changes in wave patterns. In addition increased moorage-spaces
-will mean increased,boat usage-on the lake- This,will be accompanied
by an increased amount of waste producta, from the cooling systems,
bilges, etc., being dumped into the water. Both the change.in wave,
patterns and the@increased waste.,products cart affect the,:usefu1-.life_,
.:Of-buildingsland-other structures., The costs of the project-must in-m
-clude:the value of output foregone because of the damage,. or the.costs
of protection made necessary,-whidhever is smaller.

3.4.2 Monetization of Value of Materials Soiling
Resulting from Exposure to Pollutants

Costs of increased supplies and manpower in maintenance of households
and the environment over those which would be required at a lower
level of pollutant are calculated.

This method probably does not measure the total impact of materials
soiling, i.e., the aesthetic effect of viewing soiled properties whose
owners will not or cannot afford.to pay the cost of maintenance.

The-same points discussed..in subsectidn,3.4.1,apply he re as well.

3.4.3 Monetization of Effects on Vegetation

Economic estimates are made of losses caused by specific pollutants on
yieldi quality, and marketability of agricultural products. Loss
estimates usually concern visible, direct effects on values of agricul-
turalt commercial horticultural and commercial forest crops.

Effects on wild and domestic vegetation are typically neglected because
of difficulties inherent in measurement of the quantities affected.
(Resource: Stanford Research Institute, 1970.)

It is expected that there will.be minimal disruption of agriculture
Aue to lakefront development, but the points raised here apply equally
well.to,.fauna,. most notably fish.. It is important to separate.fish
caught for.,-ebomercial purposes from that obtained by recreationalists.i.
The loss of:the%former can be,measured by,the decrease in profits-of
theconmercial fisherman after they have had a chance to adjust to the
--biological disruption@of the.lakeand if necessary, to other lines-of
work. To do.this it will'be.necessary to measure the current profits,
..of.-commercial fisherman and,then,determine howthe disruption.caused
by-lakefront development will affect the quality and quantity of their
catch and their-revenues. It:will also be necessary to-determine how
the costs of.operation will..be affected, if at.all. Using this infor-
mation the decrease in profits.can-be determined,. The measurement of
recreationaltishery losseaAs.more difficult to determine.

3.4.4 Effects on Residential Property Values,

The relationship between property,value and various.factors related to
influences,on that.value.(percentage of homes in thearea recently
built,, houses'per-mile, accessibility. to business'districtj etci) is
established.through standardmultiple@regression analysis. One of the.
contributory,factors is.assumed.to-be the mean concentration value
of some@pollutant. From this,, an estimate can be calculated for the
monetary property value,of an incremental change in.pollutant.

Given that conventional sources of statistical inaccuracy are controlled-
for, a question arises. "What is being measured?" - -at best,..it is
the,economic value of.the purchaser's perception of deviations from
environmental quality. Deviations which@were not manifested before
the purchase, or which senses cannot perceive, cannot be measured by
this method. Thus, the value of the benefits is understated. (Resource:
Plager et al., 1976.)

Environmental quality parameters that can be important in these analyses
include turbidity of the water, noise and congestion of people using a
park near residences and,.other negative influences; as well as such
positive aspects-such as:increased flood:and erosion protection and
easy access,to recreational facilities. Thiirefore.lakefront development
can,have either a positive or@a negative effect-on land values, depending
upon which influences are more powerful.

This is a useful.,measurement procedure if used-andAnterpreted correctly.
Several.points should be made, however. -First, these are studies
that novice analysts should not undertake. They requirela qualified
ecoInometrician who can properly set up and run the multi-variate.re-
gressions. It is easy to misstate the equation by omitting relevant
variables, placing too much emphasis on the difference between pollutant
levels., It is.also possible that there will be a great deal of,multi-
collinearity present which will require extra work on data selection.

Alsol.if certain things such as increased housing protection activities
are included, using.this measure can result in double counting. It is
obvious that house protection activities are.part of the reason for
decreased property values in-high pollution areas; therefore, counting
them as a financial loss:and then again as part of a reduction-An
property value will result in an overestimate.-of the cost.. In those
instances where most of the disruption..results in the-necessity of
protective actiou,,.the most useful way to identify this cost will be:-
to measureAirectly the protection costs-involved and ignore the---
econometric analysis. Altliough.thia.method.-may underestimate.the true
-cost (i.e., property values,may-fall.further because the protective
devices are not completely-effective), it ca"n-provide reliable lower
bound estimates of this cost.. It will also mean that the services of
an econometrician will not be required.

3.4.5 Willingness to Pay Survey

A representative sample of the population islasked what-they would-be
willing to pay for a particular (descriptive) increase in-environmental
quality. An alternative-is polling &representative sample to discover
willing.to pay to forestall environmental-deterior--
ation in some respect.

The-pattern of responses in such a survey may be biased by many7factors
such as respondents' ignorance of pollution costa. Extreme care-must
be exercised in formulation and administration of the questionnaire
and in interpretation of results. (Resource: Fischer, 1975w)

The section briefly covers a method that can.be more useful than in-
dicated. The problem is getting,people' to reveal their true feelings.
They will most likely underestimate their willingness to pay if they
believe that as a result of the questionnaire they may be taxed to pay
for the services they use. If they feel they will not be taxed and
the results could have an effect on future provision of services, they
may overestimate:it. It is hard to predict which of these biases will
be stronger. To correct for this it is necessary to phrase the questions
properly and attempt to ask-for the same.information in dif lerent ways
in different parts of the questionnaire. Hammack and Brown provide a
useful illustration:in their study of the value of waterfowl and wet
lands. In one part of their questionnaire, they ask the respondent to
estimate the smallest amount he would.take in exchange for his right
to hunt waterfowl for a season. Later they ask by how much his costs
would have to increase before he would stop hunting voluntarily.
Both answers give estimates of willingness to pay. These estimates
were then regressed against such'things as income, seasons hunted,-

1
Judd Ham-ack and Gardner Mallard Brown, Jr., Waterfowl and Wetlands:
Toward Bioeconomic Analysis, Resources for the Future, Inc., Washington,
D.C., 1974.

success rates, etc. The regression equation provides a relationship
between willingness to pay and various attributes of a given population.

It is beyond,,.the scope,of this report to"g-f7e a detailed description
of how to construct a survey instrument that will obtain unbiased in-
formation., Adequam-treatment of.the topic would require a manual as
large as thelpresent volume. The Hammack andBrown illustration does
show one way to approach the problem. It is recommended that anyone
attemptingto try this.technique-on his own read very carefully the
H-g-mack. and grown@book, as wellas,some of the references contained
therein to become more aware of the theory and.methods.involved. Other-
useful sources may be found in the reference section of the report;
especiallyuseful is the.National Academy of �ciences report, Assessing
Demand for-Outdoor Recreation. It should be-remembered that this,
technique is very exacting and expensive,.,and in most cases should
only be used.1with-the help of &-qualified individual.

3.4.6 Impact on Recreation Site Demand as
a Function of Travel.Cost

The recreation site becomeathe focus of concentric geographic zones
of visitor origin, defined.by.round tripttravel cost between zone and
site. The,frequency@of visitor-origin from each zone,is,estimated by
sampling-visitors at the,site. This frequency and population data are
used,to calculate a;visitation rate in days:per capita (demand) for.
each zone,of origin. Using socioeconomic data gathered for each zone
on factors such as average income, median educational attainment, etc.,,
and travel cost, a regression analysis is used'to, test the hypothesis
that travel cost@partially determines visitation rate. Once the
functional relationship between travel costand visitation rate has
been determined., it is-assumed that changes,in admissions cost.would
have an-effect on the-visitation rate equivalent to the travel cost
function. The cumulative behavior of the zones' demand functions for
the recreation site is then estimated by calculating levels of visita-
tion rate for travel cost, plus incremental changes-in admission cost.
This relationship is then used to monetize the willingness of visitors
to pay with:the input of estimated changes in-demand for a site due to
changes in some environmental quality criterion, such as fish caught
per angler trip or level of boating activity.

The relationship between environmental quality-and recreation site use
is generally quite difficult to estimate with confidence. Additionally,
assumptions must be made in applying the method as formulated which
placelimits on its range of application. The primary purpose of the
trip is assumed to be the visit to the sit64 Perhaps an even more
limiting.assumption.is that there are no close,substitutes for the
site in the area. (Resource: Clawson and Knetsch,-.1966; Reiling,
Gibbs and Stoerener'. 1973.)

An empirical study-which addresses problems arising from the latter
-assumption is Burt and Brewer, 1971.

Perhaps a better grasp of the method can be obtained by'reviewing a 1
.very-simple,hypothetical example aepresented-by Knetsch'and, '.Davis-.-
assume a free recreation or park area at varying distances.-from-three
centers-of population given in the table below.

Visits to a-Hypothetical Recreation Area

Cost of Visits, Visits/1,000.
citV_ Population Visit Made Population

A 1,000 $1.00 400. 400
B 2,000 3.00 400 200
C 4,000 -.4.00 400 100

@The cost of visiting the area is@of major concern and-would include-.-
such items@as transportati=,-lodging, and.food cost above,-those@in,-
curred if the trip@were not-made, Each-cost'would vary-with--the-dis.*
.-tance from the park to-the:city involved. Consequently---; the-number of
visits, or rather the rate,of visits per unit total population of each
cityj would also vary.

The visits per unit of population, in this case per thousand population,
may then be plotted against-the cost per visit. A line drawn through
the three points of such a plot would have the relationship given by
the equation of C = 5 - V, or perhaps more conveniently V = 5 - C,
where C is the cost of a visit and V is the rate of visits in hundreds
per thousand population. This information is taken directly from the
tabulation of consumer behavior. The linear relationship assumed here
is for convenience. Actual data,may very well show, for example, that
$1.00 change in cost might have only a slight effect-on visit rate where
the visit is already high in-cost, and a large,.effect on low-cost visits.

The construction..of,a demand curve of the recreation area, relating
number of visits to varying cost, involves a second step. Essentially,
it derives the demand curve from the equation relating visit rates to
cost, by relating visit rates of each zone to simulated increases in
cost and-multiplying by the relative populations in each zone. Thus we
might first assume 4 price of $1.00, which is an added cost of $1.00
for visits to the area from each of the three different centers used
in our-hypothetical example. This would.have the expected result of
reducing the number of visitors coming from each of the centers. The

"'Comparison of Methods for Recreation.Evaluation," by Jack L. Knetsch
and Robert K. Davis, from Water Research, edited by Allen V. Kneese
and Stephen C. Smith Copyright 1966 by The Johns Hopkins Press for
'Resources for the Fu@ure, Inc.
28

expected reduction is estimated from the visit-cost relationship. The
total visits suggested by these calculations for different prices or
differing added cost are given as:

Price Quantity
(added cost) (total visits)

$0.00 1,200
1.00 500
2.00 200
3.00 100
4.00 0

These results may then be taken as.the demand curve relating price to
visits to the recreation area. While.this analysis takes visits as a
simple function of cost, in principle there is no difficulty in ex-
tendi ng the analysis to.other factors important in recreation demand,
.such as'.alternative sites available,..the inherent attractiveness of
the,area in question or.at least its characteristics in,this regard,
and.possibly even some mea Isure of congestion.

This.method is most useful,where there are.distinctive.travel and cost
patterns for different users,%and, therefore, is often not useful for
estimating the value for urban recreational sites, especially if most
of the users live within walking distance.. If a significant number of
the users of improved lakefront recreational sites mustuse their car
or some form of public transportation and if they canbe broken into
.meaningful categories by travel cost, this method can be used.

3.4.7 Reduction in Cost of Avoidance

This method usually involves surveys of private expeditures on items
and actions intended as a defense against environmental pollutants.
Consideration might be given to purchase of bottled drinking water,
air conditioning equipment, outlays for sound deadening, etc. Care
must be taken to separate environmental effects caused by human activity
from those naturally occurring, over which little or no control may be
exercised. Since most defensive expenditures are likely to be less
than 100 percent effective, the true benefits of improvement will
usually be underestimated. (Resource: Barrett and-Waddell, 1973.)

3.4.8 Reduction in Cost of Treatment

When an environmental resource can be considereda public good or as a
factor input to a commercial prodtr=ion process, some.expenditures for
treatment of, a.pollutant may be required before.the resource is usable
(i.e., water treatment). An improvement.inl-environmental quality can
then be monetized as the resultant amount of reduction in treatment
cost., Other interactive effects of the process and result of the
environmental improvement should be noted and accounted for. Other

production costs and consumer reactions could substantially'affect the
analysis. (Resource: Federal Water Pollution Control Administration,
1966, pp. 71-74.)

A discussion of the intricacies of this and other production cost
savings and producer/consumer surplus oriented analyses can be found
in Freeman, 1975.

These two methods will likely be very relevant for lakefront,develop-
ment, but in the cost rather than the benefit column. The construction
of parks may have a serious effect on the supply of drinking water and
the increase in cost to avoid this or to treat the water, must be
measured and included as a cost of the project.

3.4.9 Damages Awarded as a Result of Litigation

A survey is conducted of monetary awards and the reasons for the awards
in an effort to discover the dollar value of an environmental impact
as decided in the courts of the area under consideration. An indicator
is also obtained of the-frequency of instances in which injured parties
in the area seek redress in the courts. With this information, monetary
awards can be defined as a function of pollutant level.

Resi-
Difficulties arise when an attempt is made to apply this method.
dents of urban areas appear less likely to seek private legal action
than those in rural areas in the case of air (and perhaps other types
of) pollutants. It has been suggested that urbanites are conditioned
to relatively higher concentrations of some pollutants, or that metro-
politan congestion masks the identity of specific polluters. At any
rate, the small number of cases to consider may cause the method to
falter from lack of data. (Resource: Havighurst, 1969.)

This method has a lot of problems, the most important of which is that
there is no reason to believe that courts use willingness to pay as a
criterion in assessing damages. Unless the judge or jury are clair-
voyant, they must make an estimate in much the same way as discussed
here if they are going to be correct. If they do, then it would be
more useful to fall back on the original estimates. If they do not,
the values assessed really have no meaning as far as benefit-cost
analysis is concerned. Unless it is clear that willingness to pay is
the standard used by the court, this method should not be used except
for illustrative purposes.

Other ways of measuring benefits not mentioned in the IIEQ study
follow.

3.4.10 Value of Time

Certain lakefront redevelopment plans may necessitate a change in
.traffic patterns and the change in.commuting time should be inciluded
in the analysis for accuracy. The benefit of decreases in travel time
(or cost in the case of increases) is what people.are willing to pay
for it. There are two ways of looking at this. The first is to meas-
ure the net value of goods and services that can be produced in the
time that is saved. The second is to attempt to measure the valuations
that individuals place on time. The first is subject to many.pitfalls,
the strongest of which as it relates to types of projects we are deal-
ing with, is that it is unlikely that any extra output could in fact
be produced in the amounts of time to be saved. The people affected
may all be on leisure timei.and even those who-are going to work will
not be able to convert the extratime into output because of union
contracts and other institutional restrictions.

The second approach can..be more useful'and some of the particular methods
are quite simple to handle. The basis for most of them are the actions
of individuals-who have a choice of several means of transportation where
they.can reveal their:preference for,time,savings versus extra cost.
By noting the differences in travel time and costs between a chosen
method and a relevant alternative for a group of individuals making a
similar trip, (i.e., analternative where all factors.such as comfort,
service, etc. are reasonably,similar) it is possible to estimate the
trade-off between the two. For example, one-very simple method is to
plot these differences on a standard four quadrant diagram. See Figure
3-2. Each point represents the combination of cost and time difference
for one person. Those in the fourth quadrant represent individuals who
are willing to give up money to save time, while the opposite holds
true for points in the second quadrant. Individuals represented in
the third quadrant are somewhat irrelevant to the analysis because the
method they use is the.cheapest and fastest available. It is unlikely
that there will be any points in the first quadrant because that would
represent a travel mode that is more expensive and yet slower than any
alternative.

A rough but workable.estimate of the relationship between time and
willingness to pay can.be found by drawing a straight line through the
origin such'that the minimum possible number of points lie to the
northeast of it. The slope of this line can then be used as an estimate
of how people value travel time. In our hypothetical example, the
trade-off is $.35 for every 15 minutes. The justification price is
that it will minimize the:number of people who are made to appear irra-
tional. Note that at point B, the individual gives up something,-_-@,Iess
than 15 minutes in order to save $.35, and given the $.35 for 15 minute
price he has made a correct choice. Individual A, however, gives up
more than 15 minutes to save the $.35 and given the price he should
have chosen the alternative method. Obviously, this is not a perfect

metho d but it can provide a useful approximation with little formal
analysis. Using this method-of British data, it was estimated that the
trade-off was somewhere between 30 percent and 50 percent of the wage
rate for a comparable time period.

3.5 COMPARISON OF MRASUREMENT METHODS.

In all a total of 10 methods of estimating willingness to pay for certain
types of benefits have been presented. It will prove useful to compare
and contrast their relative usefulness. This can be done by combining
them into several different types of categories. To begin with, sub-
section 3.3.9 would be best ignored. For the reasons stated, the only
time it will be of any use is when professional social scientists have
been used by the court to ascertain willingness to pay. And if this
is.the@'case, this method should be more properly called a literature
review where all research done in the area,, not just court cases, is
studie&to determine if appropriate measures have been obtained.

Subsections 3.4.4, 3.4.5 and 3.4..6 and the value-of-time method are
quite useful for their stated purpose but will in general require
professional assistance. Subsection 3.4.6, the so-called travel cost
method for estimating recreation benefits is a possible exception if
the analyst is familiar with basic regression techniques'so that he
can find the relationship between visits per thousand population and
travel cost.

The problem of measuring the value of time as discussed can be a very
useful method. For all practical purposes it may.not be necessary to
use it, because the results of many different studies have been quite
similar In almost all cases where this method was used, it was found
that the.trade-off value of money for time was between 30 percent and
5.0 percent of the hourly wage rate. Unless there is reason to believe
that individuals in Chicago who will be affected by the lakefront re-
development are significantly different than people studied in other
research, it makes sense to use these figures. That is, if one of the
direct outputs of a lakefront development project is 4,000 total hours
of decreased travel time per year, then a suitable estimate of the
benefits would be 30 percent to 50 percent of the average wage rate
times 4,000.

The other methods can be undertaken by the diligent novice, keeping in
mind that the object of the research is to find the willingness to pay
for direct outputs of the project. This provides the basic framework
for proceeding with any of these methods. The exact use of each method
will vary with every project and with each case within a project, so
it will not be possible to show a step by step procedure, but some
general guidelines are ppSS4ble. All of the methods are similar in
that they can be broken into two steps. First, it is necessary to
specify the exact effects of the project and second, determine the cost
savings or actual expenditures associated with them. For example,

3 5'

FIGURE 2 V A L U E 0 F T I M E

consider the case of materials soiling or deterioration, or of vegeta-
tion destruction.. It is necessary to determine (1) how much soiling
or deterioration will be prevented (or caused) and (2) how much money
will be saved (or spen-tz:@,,as a result.. A conscientious novice, familiar
with the project, should have no trouble in digging up enough data
from the industries involved to come up with very reliable estimates of
both the physicalamounts and the dollar savings.

It should be stressed again the appropriate role these methods play
in a benefit-cost analysis. Willingness to pay can be measured by the
market price in manyinstances. When there is a non-marginal change
in output, it will be necessary to have some idea of the relevant por-
tion of the demand curve in orderto determine consumers' surplus.
For much of the work this is all that is necessary. In some cases,
many of the outputs.of a project will-not be sold on a market. In
these cases, other ways of determining willingness to pay must be used.
This is the purpose of the methods described in this section.

SECTION 4

PROBLEMS OF MEASURING COST

4.1 GENERAL

The costs of a project as far as benefit-cost analyses are concerned
are the benefits of the foregone'alternatives. Therefore the-discussion
of benefits in Section 3 applies to costs if interpreted correctly.
There are several special problems areas with the,measurement of costs,
howev.er,.that merit:special attention. Three of these, unemployment,
taxes and subsidies, and monopoly,, will be discussed below.

4.2 UNEMPLOYMENT

If'a project will employ resourcesotherwise unemployed, the social
cost of doing:so is zero. The resources were producing nothing and so
there-are no losses from transferring them to the project. For purposes
of lakefront redevelopment, labor,and perhaps some shorefront land may
be unemployed and as,such must be given careful-consideration.

The.project may also,employ productive factors.that would have other-
wise been under-employed, (rather than unemployed). The social cost of
utilizing under-employed resources is not zerot, but equal to.the value
of.output2they would have.produced in.their alternative jobs. Unfortu-
nately it will.generally be very difficult to determine the,extent to
which ; productive factor is under-employed, so it is recommended
that this particular distinction be ignored, unless the availability of
appropriate data warrants otherwise.

Several points should be made clear. The actual financial cost of
hiring or purchasing unemployed resources will definitely not be zero,
only the social cost of using them. This is an important difference.
between a benefit-cost analysis and the financial statement of a project.
The former is a decision-making tool used to decide whether the project
should be built or not, while the latter provides information on what
the dollar cost of the project will be.

Another important point is the reason for the use of the resource. If
workers are unemployed because of.the national economic situation, that
is.one thing, but if a piece of land sits idle-because of local govern-
ment-decree, that is quite another.., In the latter case it should be
emphasized that the zero opportunity cost is-due entirely to the govern-
ment restriction. If the'same governmental organization is going to
make the ultimate decision on the project, they should be aware of the
actual opportunity cost if their restriction were lifted.

A final point is that it is doubtful that there is either 100 percent
employment or unemployment of any labor type., so appropriate percentages

will have to be applied to the labor bill;after checking various un-
employment rates. Also,.if construction is to take many years, care
must be taken@to correct for predicted changes in employment rates
over time.

4.3 TAXES AND SUBSIDIES

When measuring the cost of inputs it is important to consider possible
distortions caused by taxes or subsidies. For example, if there is a
15-percent excise tax on a certain item, then the purchase price in-
cluding tax cannot be used 'as a proper measure of opportunity cost. The
tax is merely a-transfer from the consumer to the government; it is not
.an actual cost of production. Similarly, if an input is.produced in
markets where there are government subsidies, the market price will be,
an.underestimate of the actual cost of production., If these types of
items are a significant part of financial cost, it should be adjusted
appropriately to obtain a measure of social opportunity cost. In cases
where taxes are directly related.to the provision of government services
that are inputs,to the project, they should be treated as normal costs.,
For example, if fees for.construction permits are used to pay.city
inspectors for their work on development, then.such fees should be
treated as a social cost of the project.

.Along the same line, if the Federal governmen t provides a subsidy in
the construction of a project in an area, this does not lower the actual
cost of the project. It merely lowers the cost that the local govern-
Pent undertaking the project has to pay. To be formally correct, the
subsidy should be ignored in benefit-cost analysis, unless there are
special reasons why a,narrow local government accounting stance should
be used.

4.4 MONOPOLY

,In as3lMilar vein, any items produced in monopolistic markets have a
market price that is higher than the true social cost of production.
Imperfect competition will, in general, lead to market prices being
greater than themarginal social cost of production as long as the
imperfectly competitive firms seek to maximize profits. (This conclusion
may not hold for firms seeking to attain some other objective, such as
maximizing market share subject to earning some minimal profit.) The
profit-maximizing monopolist's price includes a monopoly profit as well
as the cost of inputs. To be precisely correct in measuring the social
cost of inports, their market price should be corrected for the monopoly
profit This is a difficult task. In most cases it can safely be
ignored. In instances where a large part of the financial cost is for
inputs that are produced by these types of producers, efforts should be
made to include this information in the analysis.

SECTION 5

COMPARISON OF BENEFITS AND COSTS
IN DIFFERENT PERIODS OF TIME,

It is obvious that in most)cases the benefits and costs of a project
will be provided over a period of years. The existence of a positive
rate of interest is evidence of the fact that consumption now is worth
more than consumption in the future; else, why is it necessary to pay a
premium in order to get consumers to give up the right to constme,in the
present? This being the case, it is necessary to formulate a method
whereby benefits and.costs in different periods can be.made comparable.
The proper way of doing this is by discounting. This section will des-
cribe the process of discounting:and provide some analysis on selecting
the discount rate.

If M0 represents an amount of money (say, $1,000) and i is the interest
ratel(for ins.tance,.8:percent), then by loaning it out it is possible to.
obtain M 1a year from now (in this@example, $1,080) where:
M1. = M0 (1 + i)

Similarly, if it is loaned out for two years it is possible to obtain
M2 in two.years, where:
M = M (1 + i) = M (1 + i)(1 + i) M (I + i) 2
2 1 0 0

In general terms, at the end of n.years, it is possible to obtain M
where: n
Mn = M 0 (1 + )n
M can be called the future value in,n years of M with an interest rate
0
of i. By simple algebra it follows that the present value of M n (a
specified amount n years in the future) can be represented by

M
n
Mo 0 +i)@4

The benefits.and costs in different years can be made comparable by
transforming.them.all to present values. Because they will then be
measured in the same units.,it is:permissible to add them to:obtain a
total.present value of all benefits or costs. In order to obtain a
single measure of value of a project that provides benefits B B
B B n; in years 1, 2, 3,...n,, respectively the following eq'uation
can be used.

B B B B
Total discounted 1 2 3 n
present value of - + 2 + - 3 +
benefits I + i (I + i) (1 + i)

The analysis is exactly analogous for compiring costs o,f.a project that
occur in different years.
Total discounted CI C2 C n
present value of:= - + 2+
costs I + 1 (1 + i) (I + i),

To illustrate the use of these formulas, consider the following example.
Suppose that a-city project is expected to yield annual benefits over a
three-year period worth $100,000 in the first year, $50,000 in the sec-
ond, and $200,000 in the third. The anticipated costs of the project
are $175,000 in the first year, $140,000 in the second year, and zero
thereafter.

What is the present value of these benefits and costs?. According to the
above formula, the discounted present values are (assuming a discount
rate of 10 percent):

Discounted benefit s.= $282,400 $3.00,000 + $50,000 + $200,000
(1.10) (1.10) 2 (1.10) 3

Discounted costs $175,000 + $140,000
$274,715 -1-1
io) (1.10)2

Using a discount rate of 10 percent the discounted benefits in this ex-
ample exceed the discounted costs, suggesting that this particular proj-
ect is worthwhile on economic efficiency grounds.

The importance of the discount rate is easily illustrated within the
context of this example. Suppose that we have discounted the stream of
benefits and costs with a 20 percent discount rate rather than a 10 per-
cent rate. With this one change, the discounted benefits would fall to
$233,800, whereas the discounted costs fall only to $242,935. At this
higher discount rate, it now appears that this hypothetical project is
no longer@desirable on:economic efficiency grounds. Clearly, whether
this particular project enhances economic efficiency depends upon the
opportunity cost of the funds and resources-committed to it. This op-
portunity cost is reflected, by the discount rate.

A further illustration of the importance of the discount rate is pro-
vided by Table 5-1. This table indicates the present value of $1,000
discounted over various time periods at selected discount rates. When
the discount rate is zero,.the present value and future values.are iden-
tical,'since a discount rate of zero implies an indifference between
equal amounts in the presentand the future: $1,000 now is the same as

$1,000 30 years from.now. However, as the discount rate is increased,
future effects begin to diminish in relative importance. For example,
$1,000 to be.received 30 years.from today is worth only about $2 today
when discounted at 15 percentt

Table 5-1

Present Value of $1,000 for Various
Time Periods and Discount Rates.

Number of Years.Until Discount Rate
$1,000 Benefit Received 0%. 3% 7% 10% 1
5%

1 $1,000 $971 $935 $909, $870
5 1,000 863 713 621 497
10 1,000 508 386 .247
20 1,000 554 258 140 61
30 1,000 412 131 57 2

While there is controversy,over what constitutes the appropriate proce-
dure for calculating a,d.iscount rate, in many instances this controversy
need.not concern the benefit-cost analyst. It will often be,the case
that,the exact value.selected for the discount rate.will'not affect the
outcome of the benefit_cost.study@. For example,some projects offer
such a high rate of return that they would appear desirable according to
benefit-cost criteria regardless of the discount rate used. However,
,the exact value ofthe discount,rate is important in.situations involv-
ing mutually exclusive projects, budget-constrained agencies, or projects
that appear efficient f.or'some discount rates, but inefficient for others.

The conventional method for identifying those instances in which the
value of the discount rateds important is to perform a sensitivity
analysis, that is, to calculate the present value of benefits and costs
for somealternative discount rates. The analyst should select a re-
latively low discount.rate (e.g., 5 percent) and a relatively high rate
.(e.g., 20 percent) to test the sensitivity of the benefit-cost estimates
to the discount.rate. If. it is determined that the benefit-cost meas-
ures are relatively insensitive to changes in the discount rate, the
matter essentially ends there; however., if the study's findings are,
sensitive to the discount rate selected, the analyst must make that fact
known to policymakers.so that they can make their own judgments about
the appropriate magnitude for the discount rate.

During periods of inflation, t 'he analyst.must take care to avoid having
inflationary effects'influence the findings of the benefit-cost study.
To illustrate how inflation can affect a benefit-cost analysis and how
inflationary influences can be eliminated, consider the following simple
example.

Suppose a proposed project offers an estimated annual net benefit
stream (assuming no inflation) of $100,000 over three years for an im-
mediate outlay of,,,$249,000 -- an implicit rate-of-return of 10 percent.
Supposing further that the cost of funds to the City is 8 percent, the
proposed project would appear desirable-according to benefit-cost cri-
teria since,discounting the future benefits at 8 percent yields a present
value of benefits equal to @258,000,

Suppose, however,,that.an annual 5 percent inflation is anticipated for
the foreseeable future. If this expected inflation is incorporated into
the estimated futurebenefits, the benefit stream (rather than being
$100,000 annually) would be.$105,000, $110,000, and $116,000, respec-
tively, for-each of the three years. Obviously, discounting this bene-
fit stream at 8 percent would-yield an even higher present value of
benefits.than $258,000, apparently suggesting that anticipated inflation
enhances-the efficiency value of a project. However, this suggestion
should be rejected on the grounds that the 8 percent discount rate (as-
sumed appropriate for a non-inflationary.period) is no longer appropri-
ate in this situation.. If financial markets have fully adjusted to the
anticipated 5 percent inflation, we would expect the private opportunity
cost of funds to rise in tandem with the rate.of inflation. In the ab-
sence of inflation, a $100,000, one-year investment in the private sector
would, on the average, return $108,000 after one year (an 8 percent re-
turn). With a 5 percent inflation expected, the payoff at the end of
the one year would presumably be $113,400 ($108,000 x 1.05), for a nomi-
nal rate-of-return of 13.4 percent. Thus, a 5 percent fully anticipated
inflation can be expected to increase the private opportunity cost of
funds from 8 percent to 13.4 percent (in the present example).

Using a discount rate of 13.4 percent to calculate the present value of
the three-year time stream of inflated benefits ($105,000, $110,000, and
$116,000) yields exactly $258,'000, which is the same estimate obtained
for the non-inflationary situation.

The conclusion one can draw from this analysis is that inflation can be
appropriately handled by either of two methods: 1) estimate all future
benefits and costs in constant prices .(that is, simply assume no infla-
tion in the estimation of future effects) and use as the discount rate
an estimate of the private opportunity cost of funds in the absence of
anticipated inflation, or 2) estimate all future effects in inflated
prices and use.as the discount rate an estimate of the private oppor-
tunity cost of.funds in the presence of anticipated inflation.

Unfortunately, regardless of the approach taken, the analyst will en-
counter difficulties that are not easily resolved. The first approach
requires the analyst to make some judgment about the private opportunity
.cost of funds in.an inflationless, setting. When inflation is anticipated,
the perceived or nominal return to private investment contains an infla-
tion premium that may be very difficult to estimate. If the current rate
of inflation is used as an estimate of the inflation premium, it will

sometimes be the case that the implied "real" return to private expendi-
tures will be negative. On the other hand, if the second approach is
used,the analyst.must contend with the problem..of predicting therate
of inflati on:1,ro-r,many.years into the future.

In practice, it appears to be the.case that most,analysts, estimate future
benefi.ts,and costs in constant*prices, thatis,,they ignore inflation in
the estimation of benefits and costs. As an,approximation to the private
opportunity.cost of funds in-an inflation-free setting, it is.common
practice to arbitrarily select a "plausible" rate of return (e.g.,, 10
percent) for discounting.the,benefits and costs of public sector proj-
ects.

There is a great amount.of literature on-the proper discount rate to use
but it is.'too,long and complex to review here. In keeping with our con-
cept of opportunity cost, there is-agreement.among many economists that,
the,rate.chosen should be no lower than the rate of return to capital in
the private sector. With.existing income tax laws, it is likely that-
this rate is.somewhere around 20 percent. Because the discount rate can
be so critical,it is wise to.use.a sensitivity analysis to see how the
results vary with different rates,. say between.5.percent and 20 percent.

For city (or regional) projects with narrowly defined accounting stances,
the.rate at,which the-governmental unit can borrow.can be used as a dis-
count rate for city projects. If a city can borrow funds at, say, 8
percent, then city officials,can increase local welfare by undertaking
.all projects yielding implicit rates of return in excess'of 8 percent.
'Consequently,.the cost of borrowed funds (8.percent in this example)
would be the.appropriatelrate to use in discounting future benefits and.
costs of city projects. (For further discussion of these and related
issues,..see R. Musgrave and.P. Musgrave, Public Finance in Theory and
Practive, 2nd Edition, New York: McGraw-Hill, 1976; and E -J. Misham,
Cost-Benefit,Analysis, 2nd Edition, London: George Allen and Unwin,
1975.)

SECTION 6

ALTERNATIVE BENEFIT-COST CRITERIA

There-are a number of alternative criteria for accepting or rejecting
projects. These various criteria and their associated decision rules
(which account for efficiency effects only) are:

Criteria Decision Rule

1. Benefit-cost ratio, B/C (the Accept if B/C is greater than 1;
ratio of the present value reject if B/C is less than 1.
of benefits and costs).

2. Net present value, B-C (the Accept if B-C is greater than 0;
difference between the present
value of benefits and the
present value of costs).

3. Internal rate of return (the Accept if the internal rate of
implicit rate of return return exceeds the private oppor-
offered by the project). tunity cost of capital, otherwise
reject.

As an illustration of these criteria, consider a project costing
$249,000 and providing $100,000 in annual benefits for three years.
Assuming an 8 percent discount rate, we obtain the following:,

B/C = $258-,000/$249,000 = 1.04

B-C = $9,000

Internal rate of return = 10 percent.

All three decision criteria reveal that, on efficiency grounds, the
project should be undertaken.

If only one project is under considerations, or if there is no budgetary
constraint so that all acceptable projects can be undertaken and all of
the projects under consideration are independent of one another (ruling
out mutally exclusive projects), thenthe analyst can use either
benefit-cost ratios or net present values. Under these circumstances,
the two criteria will agree as to the efficiency implications of a
project.

Usually the internal rate of return criteria will yield results identical
to the other two criteria. Unfortunately, it is sometimes possible for

a time stream of benefits and costs to imply more than one internal
rate of return. For instance, consider a project involving an initial
cost of $100,000, net benefit of $220,000 at the end of the first year
and a net second=year cost of $120,000. This time stream will yield
internal rates of return of 0 and 20 percent, obtained by solving the
equation
-$1 + $2.2/(1 + r) - $1.2/(1 + r) 2 = 0

for r, the internal rate of return.

Suppose that the private opportunity cost of funds is to be 10 percent.
Should such a project be undertaken? Is it a good investment yielding
a return of 20 percent, or is it a poor one offering a zero return?

The best way to resolve this difficulty is to discount all future
effects back to the present, using the private opportunity cost of
funds as the discount rate. This approach yields $199,980 as the
present value of the benefits and $199,125 as the present value of the
costs, suggesting that at a 10 percent discount rate the project would
be marginally,beneficial. In fact, in this example, any discount rate
between 0 and 20 percent would reveal that the project should bender-
taken on efficiency grounds.

One might infer from this example that the two internal rates of return
provide information about the maximum and minimum discount rates (0 and
20 percent) that would yield positive discounted net benefits. Under
some circumstances such an inference would be correct, but under other
circumstances it would be seriously misleading. To illustrate, let us
simply alter the preceding example slightly, so that the time stream is
$100,000; - $220,000;-$120,000 (rather than-$100,000; +$222,000
-$120,000). This flow of benefits and costs also implies internal rates
of return of 0 and 20 percent.

However, discounting these effects with rates between 0 and 20 percent
yields discounted costs in excess of discounted benefits. Only if this
time stream of benefits and costs is discounted with rates less than
zero or greater than 20 percent would the projects appear acceptable
on efficiency grounds.

While in some instances it may prove pedagogically useful to calculate
and refer to internal (or implicit) rates of return, it is normally
advisable for the benefit-cost analyst to rely upon the less ambiguous
benefit-cost criteria, either net present values, or benefit-cost ratios.

Let us turn now to a situation involving mutally exclusive projects.
(this analysis also applies to situations in which a budget constraint
prevents an agency from undertaking all projects justified on benefit-
cost grounds.) Consider two mutually exclusive projects with the
following characteristics:

Discounted Discounted Net Present
Project benefits costs B/C value

A $ 2.0 mil. $ 1.0 mil. 2.0 $ 1.0 mil.
B $15.0 mil. $10.0 mil. 1.5 $ 5.0-mil.

Using the benefit-cost ratios criterion, project A appears better than
project B, whereas application of the net present value criterion
suggests the opposite conclusion. Since only one of these projects
can be undertaken (by assumption), which project should it be? This
.question is easily answered as soon as the discounted costs and benefits
of the projects are made comparable.

If project A is undertaken, the agency would have an additional $9.'0
.million (the difference between the cost of projects B and A) to invest
in other projects or perhaps to return to the taxpayers. The issue of
whether.project A is preferred to B on efficiency grounds depends on
how.the agency uses the funds it would save by undertaking A rather than
B. Suppose that the $9.0 million could be employed in the next best
alternative-investment to yield discounted net benefits of $10.0 million.
We now have the data needed to make the projects comparable. Taking
into account the auxiliary $9.0 million investment if project A is
adopted, we otain the following amended set of characteristics:

Discounted Discounted Net Present
Project benefits costs B/C value.

A $12.0 mil. $10.0 mil. 1.2 $ 2.0 mil.
(amended).
B $15.0 mil. $10.0 mil. 1.5 $5.0 mil.

Both criterianow agree as to which project is more efficient. Once
the projects have been made comparable, it becomes clear that project B
in this example is the more efficient of the two.

As a general rule, the analyst should keep in mind that economic ef-
ficiency considerations in principle require an agency to expand its
funds so as to maximize discounted net benefits. While it is common
practice for agencies like the Corps of Engineers to rely upon the
benefit-costratio criterion, we-would nevertheless recommend that the
net present value criterion be employed as the basic decision rule for
benefit-cost studies. Unless the benin-fit-cost patterns of alternative
projects have been made comparable (in the fashion outlined), use of
the benefit-cost ratio criterion tends to undesirably bias the decision
in favor of low-cost projects that, while yielding a high return per
dollar expended, may fail to maximize the net benefits from an agency's
activities when mutually exclusive projects are involved.

SECTION 7

RISK AND UNCERTAINTY

7.1 INTRODUCTION

The essence of.benefit-cost analysis is the identification and measure-
ment of a program's-benefits and costs. Normally, a program's effects
are experienced over some period of time, rather than immediately, so
there are naturally some uncertainties or risks in attempting.to predict
the precise,,nature of these,effects. The measurement of benefit and
cost time streams.in,effect requires th.e,analyst to predict--perhaps
over very long,periods of time--such things as.changes in consumption
patterns,, population movements and trendsi technological, discoveries,
and perhaps even weather patterns. Estimates of the-longer-,term
benefits from@lakefront development, may be of questionable reliability
because of uncertainty over such things:as future air and water quality,
availability of@alternative recreational opportunities, or long-term
.population trends in,the area.

In this section..some alternative approaches to handling the problems
posed by risky.or uncertain..outcomes are discussed., Economists commonly
drawla distinction.between.risk,and uncertainty. ",Risk" refers to
situations in:which inf ormation.is.available regarding-the probability-
of an outcome's occurrence: ",Uncertainty" refers to situations in which
no such information is available. In this discussion, these terms will
be used interchangeably to@refer to all less-than-certain outcomes. It
should be made.clear that while the following methods provide the
analyst with some alternative means of coping with risk and uncertainty,
it is often-the case that none of these methods are explicitly employed
in benefit-cost studies. Frequently, problems of risk.or uncertainty
are handled rather subjectively with the analyst relying more upon
"Judgment" than@any formal technique.

7.2 CUT-OFF PERIOD

One approach to dealing-with uncertainty is.to adopt some arbitrary
cutoff (or payback) period'. This is a time period past which all pro-
ject effects are simply ignored. In the case of extremely risky pro-
jects the cutoff period might be as short as two or three years; in
other cases it might be as long as 30 to 50 years. This strategy would
result in the adoption (onefficiency grounds),of only those projects
capable of.generating sufficient (discounted) benefits-prior to the
cutoff to more than',cover (discounted) project costs. This decision
rule is analogous to the "payback" criteria commonly employed by.
businessmen in judging the desirability@of investments.

Extremely short cutoff periods (e.g., two or three years) would seldom
appear justified in,evaluating public.projects. Normally, public sector

4 5

investments are simply not risky enough to warrant such a limited time
horizon. In addition, analyses conducted with short cutoff periods
ignore all information related to periods past the cutoff. Even when
there is considerable uncertainity about future outcomes, the analyst
would probably be well-advised to avoid discarding or ignoring potentially
useful information simply because it applies to periods past the cutoff.

Relatively long cutoff periods such as 50 or 100 years are more justi-
fiable. The longer the period of time involved (other things equal),
the greater is the presumed degree of uncertainity. The analyst'can
be more confident in predicting consumption or recreational patterns 5
years hence than in predicting such patterns 50 years hence. Moreover,
as noted in Section 5, discounting tends to render effects occurring so
far in the future relatively unimportant, so ignoring them seems justi-
fiable in most circumstances.

An advantage of this approach is its simplicity: it is only necessary
to select the cutoff point. The selection of any particular cutoff
point is arbitrary, however, so few guidelines can be provided to assist
the analyst in this particular task. Using cutoff periods shorter than
30 or 40 years is discouraged to avoid discarding potentially useful
information.

7.3 DISCOUNT RATE ADJUSTMENTS

An alternative approach to coping with the problem of.uncertainity
about the magnitudes of future benefits or costs involves arbitrary
adjustments to the discount rate (that is, the opportunity cost of funds).
In order to err on the conservative side, if this approach is used, the
adjustments should entail increases in the discount rate .Ised to discount
future costs. (Erring on the conservative side is desirable if policy-
makers are risk-averse; if they are risk-neutral or risk-takers, the
conservative adjustments of this sort may not be beneficial.) This
alteration will tend to reduce the magnitude of discounted net benefits
by simultaneously reducing the magnitude of discounted benefits and in-
creasing the magnitude of discounted costs. 'Consequently, this approach
will result in fewer projects being@adopted on efficiency grounds.

This procedure is probably preferable to the adoption of a cutoff period
since it discounts, rather than completely discards, information about
future effects. Nevertheless, it too involves some arbitrary decisions.
The analyst must decide not only which portion of the time stream of
benefits and costs should be discounted with the adjusted rates, but
also what the magnitudes of the adjustments should be.

One way of reducing the arbitrariness in the selection of the adjustment
magnitude has been suggested by Arrow and Lind. They recommend examining
the rates of return being earned on similar investments in the private
sector to determine the appropriate adjustment magnitude. If private
lakefront developments were yielding a 13 percent before-tax rate of
return whereas relatively riskless private investments were offering a 10

percent before-tax rate ofreturn, then the appropriate adjustment
magnitude would be 3 percent.(the difference between.the two rates).
In accordance with these suggestions and in the context of this example,
it w6ul&iappear natural to discount future benefits from public lake-
front developments with a rate of.13 percent. (That.is,. the before-tax
rate of return being earned on similar private sector. investments.)
If the City can borrow funds at.less than the 10,.percent used in this
example, the,conclusion would be somewhat different. For instance, if
the City could borrow funds.at 8 percent, this adjustment.would suggest
that perhaps 11 percent (8 percent plus 3 percent) is the appropriate
discount@rate.

Thisapproach helps to partially resolve the problem of uncertainty
about future,benefits and,@costs; however, as it tends to.involve highly
arbitrary decisions its attractiveness is,reduced., The following is
recommended. If it is possible to obtain estimates of before-tax rates
.of return to private investments in lakefront developmentsY discount
rateiadjustments may reasonably be made. If such estimates cannot be
obtained, some other approach.to coping with risk and uncertainty should
be.employed.

7.4 EXPECTED VALUES

Another method of explicitly introducing.risk or uncertainty consider-
ations,into a benefit-cost.analysis is to treat estimated benefits and
costs as random variables,that can be described by some probability
distribution. Suppose that an analysis of historical weather patterns
suggests that (other things equal) to average annual (discounted) bene-
fits from a lakefront development project will range from-$1 million
(extremely poor weather conditions for the year) to $10million (ex-
tremely.good weather conditions). Suppose that we can identify only
four possible outcomes and the probability that any given outcome will
occur. (in practice the number of:alternative outcomes would.be much
larger.) This information could.then be summarized as follows:

Value of discounted Probability of
benefits occurrence

$ 1 million 0.3
$ 3 million 0.4
$ 5 million 0.2
$10 million 0.1

Which of these discounted.values, or which combination of values, should
be used-in calculating thIeprojectfs benefit-rcost ratio (or net present
value)? The conventional method of determining one unique value when
this type of information is available is to calculate the project's
expected value, a weighted average of the alternative.outcomes:

Expected value of benefits ($1 million x 0.3) + ($3 million x 0.4)
+ ($5 million x 0.2) + ($10 million x 0.1)
$3.5 million

Suppose that the average annual discounted cost of the flood control
project is,$3.1 million. Should the project be undertaken? An examina
tion of the probability distribution of benefits reveals a 70 percent
chance the actual average benefits in any one year will be less than
the $3.1 million average annual project cost. The project should
probably be undertaken because the expected value calculation indicates
that the average benefit from such a project will be $3.5 million.

Over a period of 10 years one could, on the average, expect $1 million
in benefits in three of them, $3 million in benefits during four of
the years, $5 million in two of-the years., and $10 million in one year
for an average annual (discounted) benefit of $3.5 million.

it is likely that each estimated annual benefit and cost may be a random
variable, suggesting that the appropriate discounting.procedure (when
information about probability distributions is available) is to trans-
late expected future benefits, E(b), and costs, E(c), into an expected
present,value, E(PV): for instance:. .
E(PV of+benefits) = E(B ) + E(B 1)+ E(B 2) + ... + E(B n)
0 -
+ r + r) 2. + r )n

One serious problem with these expected value calculations is that they
take into account only the mean of the distribution, ignoring other
potentially important characteristics of a distribution (e.g., the
variance). For instance, consider two mutually exclusive, equal-cost
projects, each offering expected (discounied) benefits of $1 million.
Should the analyst suggest that these projects are equally desirable on
efficiency grounds? Probably not, unless the benefit distributions of
the two projects are similar. Suppose the two projects (A and B) have
benefit distributions such as:

Project A Project B
Benefit Probability of Benefit Probability of
distribution occurrence distribution occurrence

$1 million 1.0 $0 0.9
$10 million 0.1

Both projects have the same expected value of $1 million, but have very
different benefit distributions. In this case, as long as policymakers
are at all risk-averse, they will prefer project A (other things equal)
with its certain gain of $1 million to project B. However, if policy-
makers tend to be risk-takers, it is no longer obvious that project A
is preferable to project B--at least from the viewpoint of the policy-
maker.

Another difficulty in using this approach is the.problem (and cost) of
discovering the nature of the necessary probability distributions.
Probabilities of some events (e.g., floods, tornadoes) may be easily
obtained from hiftarl-cal records, but the probability distributions for
many variables,(e.g., input and output prices) may be very difficult
and costly to obtain. When the relevant,probability distributions are
known, this approach is the conventional one for incorporating risk
elements into.the analysis. Thisapproach is highly recommended when
the necessary information is available.. When data on the relevant
probability distributions is not available, another technique must be
employed.

7.5 GAME THEORY- MAXIMIN STRATEGY

Lacking information aboutprobability distributions, it is sometimes.
useful to apply game theory techniques in evaluating alternative pro-
jects. 'One technique is the "maximin" strategy,, referring to "maximizing.
.the minimum.". This, approach is extremely conservative, implicitly
assuming.that the worst possible outcomes.always occur.

To illustrate, assume the analyst is evaluating three mutually ex-
clusive, equal-cost lakefront development projects (A, B, and C) that
would.generate total discounted benefits of $100 million, $120 million,
and $150.million, respectively.under the most optimistic assumptions
(regarding such things,@as weather conditions, pollution problems, etc.).
Under the most pessimistic assumptions, the projects still yield (dis-
counted) benefits in the respective amounts.of $30 million, $60 million,
and $20 million. (For simplicity only two situations are examined:.
the best and worst possible.cases.) Assuming that no information is
available regarding the probability of flood occurrence, the expected
value approach cannot be used.

The various possible outcomes can be illustrated in a simple matrix:

Best possible Worst possible
outcome outcome
Project A $100 million $30 million
Project B $120 million $60 million
Project C $150 million $20 million

The maximin strategy suggests that project B is preferable to projects
A and C because B offers a minimum benefit of $60 million versus $30
million and $20 million for projects A and C, respectively. That is,
the selection of B would maximize the minimum benefits offered by the
alternative projects-@?-r

In many instances this conservative maximin criteriacould.lead to the
rejection of the more preferable projects. Suppose the previous benefit
matrix were altered.in the following fashion:.

Best possible Worst possible
outcome outcome
Project A $800 million $59 million
Pro4ect B $120 million $60 million
J
Project C $900 million $58 million

Under these circumstances the maximin criteria still recommends project
B even though it is now almost certainly inferior to either project A
or C. The problem with the maximin strategy is essentially the same
as the problem with the cutoff period criteria: it ignores potentially
useful information, namely, all gains other than the minimum. This
approach is extremely conservative.. Reliance.on it is not recommended
unless the analyst had been directed to seek out the most risk-averse
strategy..

An obvious.advantage of the maximin approach is that it requires rela-
tively little data. The analyst only need-s estimates of net benefits
under the worst possible scenario in order to implement this approach.

7.6 GAME THEORY: MINIMAX-REGRET

An alternative game-theory approach to the maximin strategy is the
minimax-regret criteria; that is, minimizing the maximum regret-(or
loss) that might be suffered. The minimax-regret strategy more fully
utilizes the available infomation about possible outcomes than does the
maximin approach, and thus is less subject to the criticism that the
approach is too conservative and risk-averse to be useful to the benefit-
cost analyst.

The minimax-regret criteria can be illustrated using the example summa-
rized in the previous benefit matrix. Suppose that the more optimistic
outcome does occur. In that event, project C would have provided the
greatest advantage ($900 million in benefits). Had project A been
.undertaken, ratherthan project C, the foregone benefits would be $100
million (= .$900 million - $800 million). Had project B been undertaken,
rather than project C, the foregone benefits would be $780 million
(= $900 million- $120 million). If it is supposed that the worst
@possible outcome occurs, project B generated the most benefits, namely,
$60-million. If project A or C is selected, the foregone benefits from
failing to undertake project B would be $1 million and $2 million,
respectively. This information can also be presented in matrix form:

Best possible Worst possible
outcome outcome Row maximum

Project A $100 million $ 1 million $100 million
Project B $780 million $ 0 million $780,million
Project C $ 0 million $ 2 million $ 2 million

Examination of the various row maximums indicates that selection of
project A might cause $100 million in benefits to be foregone, selection
of project B entails a potential opportunity cost of $780 million. The
minimax-regret strategy would lead the analyst to recommend-,%-@,pX@,oject C,
.since the worst the policymaker could do by selecting C would be to
forego (in the event of the worst possible outcome occurring) an.extra
$2 million offered by project B.

This approach to handling uncertainity-can be useful if the number of
variables under consideration is small.. If the analyst attempts-to
use these game-theory techniques where there is uncertainty about
several variables (such as future prices, consumer tastes, technology,
and weatherpatterns), the number of different outcomes the analyst
would have to-consider increases geometrically with increases in the,
number of such variables. Consequently, these game-theory techniques
are of limited usefulness in most practical benefit-cost analyses.

7.7 SENSITIVITY ANALYSIS

Another approach to coping with risk and uncertainty (one that can
easily be used in conjunction.with any of the preceding methods) is
the application of sensitivity analysis. The analyst can be more
confident in predicting that some variable (e.g., benefit or cost) will
fall within a certain range (e.g., between $1 million to $10 million)
than in predicting a precise value from that variable. Consequently,
whenever there is considerable uncertainty about the reliability of
a predicted value, the analyst could recalculate the benefit-cost ratios
(or net present values) for some alternative values, presumably upper
and lower bound estimates of the variable in question. An advantage of
this approach is that it allows the analyst to identify those (uncertain)
estimates that are crucial to the analysis. An additional advantage is
that this approach does not require additional.data (e.g., probability
distributions). The sensitivity analyses are simply performed with
whatever data has been obtained for other aspects of the benefit-cost
study. If the sensitivity analysis reveals that even relatively Ilarge.
changes in a particular estimate do not alter the general outcome of
the study, the fact that some uncertainty may surround that.estimate
is unimportant. However, as suggested in Section 5, the amount of
information generated by an additional sensitivity analysis increases
at a geometric rate, so.this approach can quickly lose the advantages
it has to offer.

SECTION 8

ALLOWANCE FOR INTANGIBLE-EFFECTS

In some instances certain efficiency or distributional effects, though
quite important, must remain unmeasured (or measured very imperfectly)
because of incomplete data or the absence of commonly accepted meas-
urement techniques, (e.g., aesthetic benefits from a lakeshore develop-
ment). In other instances, an'environmental assessment or impact
analysis may r4veal information about the physical magnitudes of a
problem (for instance, anticipated increases in traffic congestion)
but it may not be feasible to place a dollar value on such effects.
Such unmeasured and unvalued effects we refer to as intangible bene-
fits or costs.

When confron ted with the problem of intangible benefits or costs, the
analyst should, at least, clearly identify them for policymakers. The
unmeasured effects can be far more important in some circumstances than
the quantified effef-,ts and, consequently, must not be overlooked by the
analyst or the policymaker s.

One way to incorporate intangible effects into a benefit-cost study
(other than merely identifying them), is to answer the following
,question: Does it appear likely that in any particular instance the
intangible effects 1--.ould have been large enough, had they in fact been
.quantified, to have substantially altered the findings of the benefit-
cost study? To illustrate this approach, consider a hypothetical proj-
ect yielding measured discounted benefits of $10 million for a cost of
$15 million (all costs are assumed to be accurately measured). The
measured social costs of this program clearly exceed the measured so-
cial benefits. Should the analyst recommend, on efficiency grounds,
,that the,program be rejected? The project should be rejected on these
grounds only if the intangible (unmeasured) benefits of the project are
presumed to be less than $5 million, the difference between,the meas-
ured costs and the measured benefits.

The analyst's role is not to determine whether, in this example, the
intangible benefits of the project exceed $5 million, but rather to
inform the policymaLars: 1) as.to which effects have not been meas-
ured, and 2) the magnitudes that would have to be attained by the intan-
gible effects ($5 million in this example) before they could alter the
conclusion implied by the measured effects.

This approach highlights one interesting aspect, namely, that intan-
gible effects, regardless of their true magnitude, may be relatively
unimportant to the benefit-cost analyst or the'policymaker under cer-
tain circumstances. Suppose that in the preceding example the esti-
mated discounted benefits of the project exceeded the discounted fully
measured costs, and that the only issue involved was whether to

undertake a project of agiven size. In this instance since the meas-
ured benefits exceed the costs, the fact that there may be substantial
unmeasured benefits associated with the program is irrelevant, unless
a budget constraint prevents the agency from undertakin&,,all efficiency-
enhancing projects. If the size of the project was in questiou, the
intangible effects would still be relevant, since economic efficiency
suggests that projects should be expanded until the social marginal
benefits.just equal to the project's social marginal costs.

SECTION 9

DISTRIBUTION OF BENEFITSIAND COSTS

Benefit-cost analysts tend to emphasize the economic efficiency aspects
of a project; that is, the identification and measurement of a project's
11real" benefits and costs. A decision maker may be concerned not only
with the relative magnitudes of the benefits and costs, but also with
the distribution of.the benefits and costs. In some instances the dis-
tributional considerations may be of overriding importance to a policy-
maker. Consequently, it isuseful for the analyst toindicate, to the
extent feasible, how the benefits and costs of a project are likely to
be distributed among various groups of people.

Many classifications could be used in discussing a project's distributional
effects. The more familiar categories would include income, race, sex,
age, region (or neighborhood), religion ,family size, occupation, and
educational background. These characteristics could themselves be further
divided into various subcategories.

In addition to the difficulties and costs of determining how the benefits
and costs are distributed, presentation of this type of information can
be difficult to accomplish, even for a relatively few classifications.
For instance, distributing a project's benefits and costs across 10 in-
come groups, 8 age classes, 4 racial categories, and 6 regions (or
neighborhoods) requires a table containing approximately 2,000 separate
cells., Unless,many of the cells were empty, most individuals would
probably find it very difficult to assimilate and digest that much in-
formation. Consequently, while the distributional effects may be of
considerable interest, the cost of identifying and measuring such effects
and the problems of transmitting such detailed information suggest that
the analyst should normally work with a few broad classifications.

Most policymakers will indicate a need for specific types of distributional
information, so that the analyst only need be concerned with estimating
the distribution of benefits and costs. Sometimes the analyst will be
given considerable discretion over what, if any, distributional informa-
tion is presented.

One approach to organizing some reasonably detailed information about
the distribution of a project's benefits and costs is presented in Table
9-1. This table presents information about a prcject's distributional
consequences in terms of income, age, race,-.!and residence. This table
is easily altered to allow for the inclusion of additional variables
(e.g., sex) or for the substitution of one variable for another.

ideally, distributional measures should reflect all income or wealth
changes occasioned by a project, not just those changes associated with
the "real" (efficiency),benefits and costs. Alterations in a family's
.income or wealth posi tion can be caused as easily through purely
pecuniary effects (which should be ignored in assessing a project's
efficiency) as through real effects, and, where feasible, such effects
should consequently be takenlinto account when assessing a project's
distributional impact.

Usually the benefit-cost analyst's contribution in this area is limited
to the identification.and (where feasible) measurement of a project's
distributional effects. This statement should not suggest that this
identification and measurement task.is easily accomplished. In many.
instances it is,very difficult and costly to obtain reasonably-accurate
estimates of a project's distributional effects. It is not uncommon for
distributional assessments to be limited to merely identifying whether
particular,groups are-gainers or losers, with no serious effort.made to
measurethe magnitude of a project's distributional impact.

In principle, if policymakers have revealed wh ich distributional-effects
are desirable and which are undesirable, the desirable,effects can be
treated as-a social benefit,,the.same as a project's desirable efficiency
effects. The undesirable distributional effects can be similarly.viewed
as social costs.- Theanalyst could assign weights (determined perhaps
by a legislative body or some official) to distributional gains and
losses to make them comparable across.groups of people and also with
efficiency.gains and losses. To illustrate this point, consider a
project that confers real discounted,benefits of $100,000 to group A
while imposing $200,000 in real discounted costs on group B. (For
simplicity all income or wealth changes associated with.secondary or
pecuniary effects are ignored). According to strict benefit-cost
criteria, this project should be.rejected since its net present value
(discounted benefits minus disounted costs) is less than zero. This
decision rule assumes that a dollar gain, or loss, is the same regardless
of the group.involved. Suppose, that the city's policy objectives in-.
dicate that a dollar gain (or loss) to group A should receive three times
the weight received by a similar gain (or loss) to group B. With this
information, the gains to group A would have to be revalued at $,300,000,
thereby making the project desirable on benefit-cost grounds once the
distributional gains are explicitly taken into account. The analyst
will rarely, if ever, have such explicit distributional weights available
and, thus, is likely to be restricted to simply presenting estimates of
a project's distributional effects to decision makers and letting them
assign implicit weights to the effects.

Clearly, this issue of distributional gains 'and losses can be very sensitive
politically. The extent to which this sensitivity will (or should) affect
@a benefit-cost,study will vary from project to project and from agency to
agency. Consequently, not much guidance can be offered the analyst regarding
this issue.

Table 9-1

Distribution of Net Benefits by Income, Age
Race,and Residence of Beneficiary

Income Class
Area of Residence $0 $4,999 $5,000 $14,000 $15,000 and Above
and Age Group White Nonwhite White Nonwhite White Nonwhite

Chicago

0 to 18 years
19 to 64 years
65 years and above

Chicago Suburbs

0 to 18 years
19 to 64 years
65 years and above

Illinois (other than
Chicago and suburbs)

0 to 18 years
19 to 64 years
65 years and above

United States (other
than Illinois)

0 to 18 years
19 to 64 years
65 years and above

SECTION 10

BENEFIT-COST ANALYSIS AND LAKEFRONT DEVELOPMENT: AN ILLUSTRATION

10.1 INTRODUCTION

This section discusses the application of benefit-cost principles and
writing a benefit-cost report within the framework of a lakefront
development project. The form of the benefit-cost report and the presen-
tation of the results are very important since this report serves as
the primary, if not the only means:of conveying a study's findings to
policymakers. Care must be taken in preparing the benefit-cost report,
so that valuable information or crucial assumptions to the study are
not overlooked or misunderstood by decision-makers. The format dis-
cussed below is a useful one for conveying the relevant information
generated by the benefit-cost study.

In this section steps will be outlined that should be taken in a benefit-
cost analysis and guidelines provided for performing a benefit-cost
study of the development of Chicago's lakefront. These guidelines will
provide City employees and/or officials with the information they need
to perform at least certain parts of a benefit-cost study of the develop-
ment of the lakefront. These guidelines are not substitutes for the
judgment and skills of an economist trained in the uses of benefit-
cost principles. Also indicated are which aspects of a benefit-cost
study of Chicago's lakefront development should be undertaken in con-
sultation with well-trained benefit-cost analysts.

The approach presented here is illustrative, and individual analysts
may choose to modify it somewhat to meet their own needs. The necessary
elements of a benefit@cost report will be outlined in general terms,
(refer back to Section 2) and the recommended approach within the context
of a hypothetical lakefront development study will be presented.

10.2 PROGRAM OBJECTIVES AND ACCOUNTING STANCE

As noted in Section 2, the first step in writing a benefit-cost report
is -to identify the objectives of the program being analyzed and relate
those objectives to the accounting stance employed in the study. The
relationship between the program objectives and a study's accounting
stance becomes extremely important if there is some divergence between
the two. For instance, in the case of Chicago's lakefront development
project, presumably city officials (and local taxpayers) are primarily
interested in whether the local benefits of the project exceed the local
costs, and perhaps only secondarily interested in the project's regional
or nationwide efficiency implications. Thus, the analyst should clearly
define whether the study's accounting stance is local, regional, or

nationwide in nature. (It is indicated in more detail below how changes
in the accounting stance can affect the benefit-cost estimates; also
see Section 2.)

It will not always be possible to identify with certainty the project
objectives because policymakers themselves may not have clearly identi-
fied a project's objectives, or those objectives may be seen differently
by different policymakers. With respect to Chicago's lakefront develop-
ment project, the objectives appear to be clearly stated (see the Lake-
front Plan of Chicago). Even if the objectives are not clearly defined,
the analyst will always be able to, and should specify the viewpoint,
or accounting stance, taken in the benefit-cost study. The policy-
makers will, therefore, have the information necessary to judge the
relevance of the study's findings.

In this discussion it is assumed that the accounting stance will be
designed to focus only on the local (city) effects of the project. It
will be shown, however, that those effects may have to be viewed differ-
ently under alternative.accounting stances.

10.3 TAXONOMY OF (EFFICIENCY) BENEFITS AND COSTS

The next step in the construction of a benefit-cost report is the
identification of the various benefits and costs that should, on ef-
ficiency grounds, be taken into account and, where feasible, quantified.
(See subsections 2.3 and 2.4 of Section 2) The purpose of this ex-
haustive taxonomy is to provide a benchmark for assessing the analyst's
success in measuring the program's benefits and costs. Not all of the
benefits and costs will beamenable to quantification, given the present
state of benefit-cost analysis. Still, it is very important to at least
identify all the benefits and costs so as to place those that are
quantifiable into perspective. Ideally, the various measures of a
project's gross benefits should capture all of those elements which, in
combination, equal the sum of all individual willingnesses to pay for
the project's advantages (representing the efficiency benefits of the
project). Similarly, the taxonomy of social costs should include all
those elements which, in combination, equal the sum of all individual
willingnesses to pay to avoid the project's disadvantages (that is,
the efficiency costs of the project).

Consider the application of these points to Chicago's proposed lakefront
development project. What are the social (economic efficiency) benefits
and costs of such a project? In addressing this question and developing
the taxonomy, first the benefits and then costs will be considered. In
the development of these taxonomies, take considerable care to avoid
any double-counting of benefits or costs. In the following paragraphs
there are specific instances where double-counting could easily occur
if careful consideration were not given to the development of these
taxonomies.

10-3.1 Taxonomy of Benefits from Chicago's Lakefront' Development

One obvious-form that the efficiency gains from lakefront development
would take is the saving associated with reduced flooding or erosion
(assuming that the development project would provide for improved flood.
or erosion protection along the shore) to the extent that flood and
erosion damage is reduced. A resource savings is realized and, the
value of these savings should be accounted for in an approximation of
a social willingness t'o pay for lakefront development. Reduced flood-
ing may result in an increase in property values along the lakefront.
The value of this benefit could be measured by estimating the reduction
in damage costs or the increase in property value. If both of these
measures are included as a benefit, the analyst will be double-counting
the benefit of reduced flooding or erosion.

Flood and erosion problems necessitate protective expend 'itures by-both
public and private lakefront property owners to protect against such.
hazards. Lakefront development that reduces the likelihood of flooding
and erosion would allow property owners to reduce such protective
expenditures and Ithereby free additional valuable resources for alterna-
tive uses. The value of any such savings should be counted as a real
social benefit of lakefront development because the lakefront property
owners would certainly have a willingness to pay@for such an advantage.

Reductions in the danger of water damage may lead to a reduction in the
need for insurance against such hazards. A lessened demand for insurance
may release resources (e.g., manpower) from the administration-of in-
surance policies and claims. The value of such freed resources in
their next best alternative use should be included as a real social
benefit. As the amount of insurance-related resources released as a
result of lakefront development may be quite small, this effect could
probably be safely overlooked in the benefit-cost study. The saving in
insurance premiums (as a result of improved flood and erosion control)
should not be included as a real social-benefit (i.e: , resource savings).
Insurance premiums and payments are financial transfers (similar to
taxes) and should not be included in an evaluation of the efficiency
effects of the project. Any savings in insurance premiums (unless they
are subsidized) should be approximately offset by reductions in the
insurance payments made to lakefront property owners, so these two
effects will more or less cancel each other.

If those who are presently running the risk of incurring flood or
erosion-related losses dislike accepting risks, they will also have
willingness to pay for lakefront development simply because it offers a
means of reducing those risks. This willingness to pay arising out of
risk-aversion should, in principle, be measured and included as a real
social benefit of lakefront development. In practice it is often not
feasible to quantify this particular benefit. Moreover, this effect
may not be of sufficient magnitude to justify the time and expense as-
sociated with gathering the data necessary to measure it.

One of the objectives of the lakefront plan is the development of
natural areas for wildlife habitats along the Chicago lakefront. The
willingness to pay of individuals for such natural habitats should be
estimated and included as a real benefit-of---the project.

An other form of real benefits from lakefront development is the increase
in recreational opportunities afforded by such a project. Individuals'
willingness to pay for enhanced opportunities for such things as
swimming, boating,,biking, and picnicking constitutes a reaLsocial
benefit of lakefront development and should be taken into account.

Lakefront development may have a beneficial impact upon employment and
income in Chicago. The project may provide more jobs, thereby directly
lowering unemployment in the city.. This direct effect is.categorized as
a reduction in the social cost of the project (see Section 10.3.2),- and
should not be counted as a social benefit in order to avoid double-
counting this effect. As,long as this direct employment effect is in-
cluded, it doesn't matter whether it is.categorized as a gain in social
benefits or a reduction in social costs; it is merely conventional to
classify this effect-as the latter.

Lakefront development may also indirectly enhance income and employment
opportunities in the area lay attracting new business.into Chicago and
adjoining communities. To the extent that the project-does indirectly
generate.additional jobs for city residents, the City will enjoy a real
benefit (not elsewhere cou-Ited) that should.be included2'in assessing
the benefits of the project Note, that these "additional jobs" repre-
sent a true social benefit only if their existence implies a lowering
of the unemployment rate (other things equal). Thus, the appropriate
procedure for measuring these indirect multiplier effects requires
that the analyst estimate (a) the number of unemployed people that will
become employed as an indirect result of the lakefront,development, and
(b) the average annual'income these people will earn. Multiplying these
two estimates together provides, an estimate of the aggregate annual
value of this particular.benefit. For a precise estimate of this in-
direct employment effect, the analyst must use either an input-output
model or an econometric model of the Chicago region. A qualified econo-
mist should be consulted on this.point if.such estimates are desired.
Less precise estimates of this employment effect could be obtained by
using "employment multipli--rs" derived in other studies.

If the accounting stance is. regional or national rather than local,
these indirect employment and income effects must be examined from a
regional or national viewpoint, respectively... The analyst must try to
determine the extent to which the project lowers the regional, or
national unemployment rate. In this connectibn,..different accounting
stances can lead to significantly different estimates of the indirect
employment effect because increased business activity and employment in
one area may develop.at tlie expense of business activity,and employment
in another area. AJocal.accounting,stance may reveal some indirect

employment benefits from the project, whereas a regional or national
accounting stance might indicate the existence of offsetting employment
and business reductions (outside the Chicago region).

10.3.2 Taxonomy of Social Costs from Chicago's, Lakefront Development

In this section efficiency costs likely to Accompany a lakefront develop-
ment project are identified. These real social costs, in principle,
must be offset against the social benefits noted previously to determine
whether the project is desirable on efficiency grounds. (See'subsection
2.4 of Section 2.)

One social cost created by lakefront development is the variety of
construction costs associated with implementing the lakefront plan.
The costs of creating such things as new beaches, offshore islands, new
marinas, tennis courts, or bikeways are all costs that must be included
as a real social cost of the project. This category includes con-
struction expenditures by both the city government (or other govern-
ments) and by private parties.

Lakefront development will also create operating and maintenance ex-
penditures (e.g., of the new parks and marinas). Also, the development
may increase operating and maintenance costs for existing facilities.
These outlays all constitute real resource expenditures and must be in-
cluded in an accounting of the social costs of lakefront,development.

Further lakefront development may entail some adverse environmental or
ecological consequences-or cause expenditures to avoid such conse-
quences. In either event, these undesirable effects would represent
real social costs that should be quantified as one of the undesirable
effects of lakefront development.

Another potential cost of lakefront development is the value of the
opportunities foregone by not employing the lakefront areas in their
next-best use (e.g., for high-rise apartments or office buildings).
The analyst must be careful, however, with this particular effect to
avoid a double counting of social costs. The market value of a piece
of land is a measure of the (discounted) benefits that can be expected
from that land. To the extent that land purchases (if any) are included
in the construction cost. category, these foregone benefits are already
included in the assessment of the project's social costs. For those'
areas acquired at less than full market value (e.g., areas already
owned by the city), the analyst must endeavor to estimate this particu-
lar social cost of development.

This completes the illustrative taxonomy of benefits and costs likely
to be caused by the lakefront development project. In an actual benefit-
cost study, the analyst may want to explain the economic rationale for
each benefit and cost in greater detail.

10.4 BENEFIT-COST MEASUREMENTS

After the taxonomy of benefits and costs has been developed, theLat)Proach
used in measuring those effects that were quantifiable should be dis-
cussed in detail (refer back to subsections 2.5, 2.6 and 2.7 of Section
2). Emphasis should be given in this section of the report to justi-
fying, highlighting the more restrictive and controversial assumptions
made in constructing the benefit-cost measures. Any divergence between
the operational measures and theoretically appropriate benefits and
costs should also be clearly indicated. For each effect measured, there
should be a separate section where the measurement techniques and data
sources are carefully explained.

An example of the type of information that might be presented in this
section of the report might be the recreational benefits that could be
expected from such a project. In connection with the construction of a
measure for this benefit, the type of information that should be pre-
sented includes:

The method used in identifying the types of people that will
benefit from the recreational opportunities.

The approach and data employed in estimating these individuals
willingnesses to pay for the types of recreational opportunities-
offered by the development plan.

The method of accounting for any anticipated future changes (e.g.,
growth) in the demand for such recreational opportunities.

The discount rate (or rates) used in translating future benefits
into present values.

The time period employed in the analysis.

In connection with the most important assumptions, the analyst should
undertake sensitivity analyses to determine how crucial a particular
assumption is to the benefit-cost estimates. A large number of crucial
assumptions will necessitate some judgment in selecting the number of
sensitivity analyses because the amount of information generated by
additional analyses increases at a geometric rate. If, for every key
assumption or variable (e.g., the discount rate), the implications of
only two alternative values,(e.g., 4 percent and 10 percent) are ana-
lyzed, the number of cells needed to display the resultant information
is 2 nwhere n is the number of key assumptions or variables examined@'.,u-
The presentation of the results of a sensitivity analysis involving 10
variables (with two values for each variable) would require over 1000
cells. If each of the 10 variables takes on three alternative values,
the number of sepawe pieces of information skyrockets to almost 60,000
(or to precisely 3 ). Sensitivity analysis can be a very powerful in-
strument, but caution must be exercised to avoid carrying its use too
far to provide comprehensible and useful information.

In the remainder of this subsection-are outlined some approaches to
operationalizing measures of the various social benefits and costs just
identified. Benefit-cost analysis is too complex and requires too many
study-specific judgmental decisions to be able to provide the detailed,
step-by-step'directions necessary for a novice to successfully under-
take a benefit-cost study. A sk illed benefit-cost analyst will probably
need to be consulted on various aspects, in order to conduct the study
correctly.

10.4.1 Measurement of Social Benefits

The.measurement of benefits associated with the development of Chicago's
lakefront will be discussed in this subsection.

10.4.1.1 Reduced Flood and Erosion Damage. The development should
reduce the amount of damage caused by flooding and soil erosion. The
reduction in such damages should be valued and included as'a real social
benefit of the project. The estimated value of this benefit involves:
(1) determining the extent of the damage without the development, and
(2) determining the extent of the anticipated damage with the develop-
ment (the with/without methodology).

Historical records should contain information useful in estimating the
extent of damage without the project. Such records should be used to
obtain an estimate of the value of the average a'nnual flood and erosion
damage under present circumstances. Since this estimate is to be used
to approximate the probable annual damage (without the development) for
many years (perhaps decades) into the future, the analyst may wish to
adjust the estimate to allow for future population growth, more intensive
use of lakefront property, rising property values, etc. Once this annual
time-stream is estimated, it must be discounted back to the present with
the appropriate discount rate (or rates). (See the previous discussion
on discounting.)

Estimates of flood and erosion damage that is likely to occur with the
project completed may be more difficult to obtain. It would be wise to
consult with marine scientists and engineers to obtain some idea of the
extent.of flood and erosion damage in the presence of the development.
If it is estimated that a 75 percent reduction in physical damage is
likely, it would be reasonable to assume that the reduction in the value
of the damage would also be 75 percent. Thus, if estimated annual
damage without the development is $2 million, and marine scientists and
engineers anticipate a 75 percent reduction in physical damage from
flooding and erosion, then $1.5 million can be used as an estimate of
the annual value of this particular benefit. These annual amounts
would have to be discounted before they could be included in the calcu-
lation of a net present value or benefit-cost ratio.)

These points must be qualified to allow for the following possibility.
The development may encourage more intensive use of the lakefront so

that in the event of a flood (less likely because of the development),
more damage occurs than would have been the case without the develop-
ment. In this instance, a 75 percent reduction (for example) in the
probability of flooding does not translate-@-rinto a 75 percent savings
in the damage caused by flooding. Floods occur less frequently, but
are more damaging when they do occur. It may not be feasible to incor-
porate this qualification in estimating the annual reduction in flood
and erosion damage because of the difficulty of predicting how the
lakefront will develop (both with and without the project). The ana-
lyst should keep this qualification in mind, though, and note that it
introduces an additional element of uncertainty into this phase of the
analysis. If this qualification cannot be implemented, the analyst
should recognize that the resultant estimated benefits from reduced
flood and erosion damage represent u2per bound estimates of benefits,
that is, they'probably overstate the true benefits of the development.

10.4.1.2 Reductions in Protective E?cpenditures.. Improved flood and
erosion control associated with the development may allow private and
public lakefront landowners to reduce the amounts spent on flood and
erosion protection. These reductions would represent a social benefit
over and above the benefit associated with reduced flooding and erosion
damage. Improved flood control may allow substantial savings in build-
ing construction costs along the lakefront, since (with the project),
the buildings could be built to withstand fewer and less powerful floods.
These cost savings constitute a real social benefit of the development.

In quantifying this benefit, the analyst should consult con.Etruction
contractors and architects to gauge the amount by which construction
costs might be lowered if flooding and erosion were less of a problem.
It might be possible to obtain measures of protective expenditures by
property owners along other shore lines, where flooding and erosion
risk is comparable to that expected along the Chicago shoreline once
the development is complete. Then such estimates could be compared to
current expenditures by Chicago lakefront property owners to gain an
idea of the possible savings that might result from the development.

10.4.1.3 Savings in Insurance Administration Costs. The extent to
which flooding and erosion reduction along Chicago's lakefront reduces
the overall cost of insurance administration on lakefront prcDerty is
likely to be negligible compared to other project benefits arl costs.
For practical measurement purposes, this particular effect can be ig-
nored, although it should certainly be included in the taxonomy of benefits.

10.4.1.4 Benefits Related to Risk-Aversion. As noted in the benefit
taxonomy, if lakefront property owners are ris-k-averse (dislike accept-
ing risks), they would have a willingness to pay for lakefront develop-
ment because it would reduce the risk of flood or erosion damage. In
principle, this benefit should be taken into account, but in practice
it is usually quite difficult to estimate the magnitude of benefits

related to risk-aversion. It is recommended that Chicago analysts
treat this particular benefit as an intangible effect, and not worry
about quantifying it.

10.4.1.5 Wildlife Habitats. Creation of natural wildlife habitats
along the lakefront is another potentially beneficial effect of the
project. In attempting to quantify this particular benefit, the ana-
lyst is confronted with approximating an individual's willingness to
pay for wildlife preservation along portions of the lakefront. This
is difficult and complex since there is no wildlife preservation market
to reveal people's willingness to pay for such a service. Opinion-
poll techniques might give a rough approximation of the value people
would place on such wildlife habitats. If this approach is chosen, it
will be necessary to consult with individuals skilled in the survey
techniques. Alternatively, it may be advisable to simply treat this
benefit as an intangible effect (see the discussion below). If city ana-
lysts choose to try to quantify thi's particular benefit, they should
consult with a qualified economist about the various avenues open to them.

10.4.1.6 Improved Recreational Opportunities. One of the main objec-
tives of the proposed lakefront development is improved recreational
opportunities for Chicago area residents. The appropriate measure of
this efficiency benefit is the residents' willingness to pay for these
opportunities. There is no commonly accepted methodology for estimating
this willingness to pay for urban recreational opportunities provided
by park-like developments. The conventional method of-approximating
willingness to pay for suburban or rural park visits involves surveying
visitors to determine what travel expenses they actually incurred to
make the visit. This information, along with personal data (age, income,
education, etc.), is then analyzed with econometric techniques to find
a rough estimate of willingness to pay. This technique is not very use-
ful for urban settings because many visitors to urban parks travel to the
urban area for a variety of reasons (e.g., shopping, business), besides
visiting the park. Consequently, it is extremely difficult to determine
what portion of an individual's total travel costs were specifically for
visiting the park or development.

One alternative is to survey individuals (potential users of the develop-
ment) to determine their willingness to pay for the recreational oppor-
tunities offered by the lakefront development. Using this approach, it
would be necessary to consult with experts in the field of opinion poll-
ing and economists to be sure the relevant data were being collected.

It is also possible to estimate willingness to pay for those components
of the development where information exists about market.price. It would
be possible to approximate willingness to pay for an additional public
swimming pool by using the fees that people presently pay at existing
(public or perhaps private) pools, along with an estimate of the number
of users of the new pool. Suppose that public pools are charging $2 per
visit and the existing pools are relatively crowded all summer. Suppose

that an additional pool would experience 10,000 visits a year. Willing-
ness to pay for this pool could be estimated to be $20,000 per year.
If the existing pools are highly congested--IT7-t-bis amount would probably
understate true willingness to pay for an additional pool. On the
other hand, if the existing pools were relatively uncongested, this
estimate would be an overstatement of true willingness to pay. There
are a number of subtleties -associated with the use of this particular
technique, so it should be used only in consultation with a qualified
economist.

These benefits could also be treated as intangible effects (see below).
This approach should only be used as a last resort since these benefits
are so important to the success of the project.

10.4.1.7 Income and Employment Effects. Employment provided directly
by the development should not be classified as a social benefit of the
project. If people employed by the development would have been employed
elsewhere without the development project, employment by the project
represents a social cost (either a construction or an operating and
maintenance cost), not a benefit. If development employment provides
jobs for people who would otherwise have been unemployed, then budgeted
construction costs or operating and maintenance costs should be reduced.

If a citywide accounting stance is adopted, a real employment and in-
come benefit arises from the development if it stimulates additional
business activity and employment in the city. Suppose that lakefront
development attracts additional visitors to Chicago, stimulating eco-
nomic activity in the area, and assume that this additional business
activity creates a net expansion of 25,000 jobs for Chicago area resi-
dents. Assuming that average income provided by these new jobs is
$10,000 before taxes, the social benefit attributable to the develop-
ment would be $250 million annually ($10,000 times 25,000). Develop-
ment-related income gains to people in existing jobs should be added
to this amount to arrive at an estimate of the aggregate employment
and income gains arising from the project.

Of course, these annual benefits would have to be discounted before they
could be used in the calculation of a net present value or benefit-cost
ratio. If a national accounting stance were used, these would not be
considered as benefits because assuming full employment, the increase
in business activity in Chicago would have to be matched by a decrease
in activity elsewhere in the nation. Estimation of these benefits re-
quires the analyst to forecast both the indirect employment effects.of
the development project and the average income.@generated by these new
jobs. Forecasting the employment and income effects is especially
difficult and requires an economist skilled in forecasting techniques
and econometric analysis.

10.4.2 Measurement of Social Costs

Presented in this subsection are some rules-of-thumb applicable to
measuring the social costs associatdd--.7-with the development of Chicago's
lakefront. '

10.4.2.1 Construction Costs. The measurement of,this particular social
cost is relatively straight orward. In principle, the figure sought is
the discounted value of the productive resources that will be used in
the construction of the development components (the marinas, parks,
bikeways, swimming pools). Construction cost estimates are readily
available, once the contractors have bid on the project.

To illustrate, if the project contractor agrees to complete the entire
development within 5 years for, say, a present value of $100 million,
then the $100 million figure would provide a good approximation of the
Teal resource costs of the construction activity. However, the ana-
lyst may be presented with estimates of construction costs.that have
not been discounted; in that case these estimates will have to be con-
verted to present value equivalents * As mentioned in the discussion of
the discount rate, the city should use the (percent) cost of borrowed
funds as the discount rate for the project.

If the benefit-cost study is being undertaken prior to the contract
bidding, the analyst may experience some difficulty in obtaining an
estimate of construction costs. In this case, the analyst might try
consulting with potential contractors to obtain the information. In
addition, construction costs for developments in other cities may provide
some useful information.

Wages of construction workers who would have otherwise been unemployed
should not be included as part of the construction costs. Such pay-
ments do not reflect thetrue opportunity cost of using unemployed
workers and should be ignored in quantifying the efficiency benefits and
costs. (See the discussion about the treatment of unemployed resources.)

10.4.2.2 Operating and Maintenance Costs. Annual operating and mainte-
nance costs will recur throughout the lif-e of the project, and the dis-
counted present value of these costs should be included in the calculation
of the project's net present value or benefit-cost ratio. These oper-
ating and maintenance costs should include wages and salaries paid to
workers engaged in the operation and/or maintenance of the facility,
and the cost of material required to maintain the equipment. Estimates
of these annual costs can be obtained from the operating and maintenance
costs of other city facilities_(e.g.,:-%pirks), from facilities in other
cities, and from private facilities (e.g., private marinas, tennis courts).

Once again, regarding the employees who would have otherwise been un-
employed, wage payments should be excluded from the estimate of annual
operating and maintenance costs. Also, it should be remembered that

annual effects must be discounted before arriving at an overall estimate
of their magnitude in present value terms.

10.4.2.3 Envir&iimental or Ecological costs. Development of the lake-
front may entail some adverse environmental or ecological consequences.
For instance, dredge and fill operations in the lake may harm the
breeding grounds of some aquatic life, thereby diminishing'the popula-
tion of a species, and perhaps even leading to its extinction. Pre-
sumably, there will be some willingness to pay the cost of avoiding
such adverse consequences, so they represent a social cost.that should
be quantified and included in the analysis. Unfortunately, it may not
be possible to attach an economic value to many environmental and
ecological effects.

The first step toward measuring these effects is an environmental assess-
ment designed to identify the physical impacts of the development. In
principle, the environmental assessment would reveal the exact nature of
the environmental and ecological consequences of the lakefront develop-
ment. After these physical effects have been identified, economic ana-
lysis can be used to evaluate some of the effects.

The valuation of environmental and ecological effects is often quite
difficult and sometimes simply infeasible. For instance, it may not
be possible to obtain a reliable estimate of the value of preserving
the breeding grounds of an aquatic species that has no commercial use.
As a result, few guidelines can be provided in this area other than to
suggest that qualified economists be consulted.

If the environmental or ecological effects impact upon a commercial
activity, it may be possible for the novice benefit-cost analyst to
derive a value for those effects, at least in part. For tnstance, if a
dredge and fill operation results in depopulation of a fish caught com-
mercially, a reasonable estimate of the social cost of this effect would
bethe resultant decline in revenues to the local fishing industry.

This estimate must be modified if the diminished catch results in
higher prices for the fish, or if the reduced fishing activity frees
productive resources (e.g., workers, boats) for other uses. To illus-
trate the first point, suppose that-the catch of a particular type of
fish falls by 100,000 pounds per year, thereby increasing the price from
$1 per pound retail to $2 per pound. If the catch reduction is valued
at the initial price of $1, we obtain $100,000 as a measure of the
annual social cost of this effect. However, if we value the reduction
at the new price of $2, we obtain $200,000 as the estimate of annual
social cost. In"this type of price-change situation, the best approxi-
mation to the true,social cost would be the average of these two estimates,
that is, the value of the catch reduction would be $150,000 annually.
(This estimation procedure assumes that the demand curve for fish is
approximately linear in the relevant range).

Prior to completion of the project it may not be possible to predict
what impact the project will have on the price of commercially caught
fish. Consequently, the analyst may have little choice other than to
use the existing market price in valuing these effects.

Next, consider the situation in which reduced fishing activity releases
productive resources for other types of productive undertakings. Suppose,
for example, that the 100,000 pound reduction in annual fish catch re-
sults in some workers moving to other types of employment, reducing
annual wage payments in this segment of the local fishing industry by
$75,000. This resource savings (valued at $75,.000 annually) must be
subtracted from the estimated social cost associated with the reduced
catch ($150,000 annually, in the preceding example) to arrive at a
measure of the net social cost.

In many cases, it will not be feasible to value environmental and eco-
logical effects. Such unquantified (in dollar terms) effects are re-
ferred,to as intangible effects. The treatment of, intangible effects
is discussed at greater length in a following subsection.

10.4.2.4 Opportunity Cost of Lakefront Land. As we noted above, another
potential cost of lakefront development is the value of the opportunities
foregone by not employing the lakefront land for other uses, such as
apartment complexes. The market value of a piece of land will normally
reflect the discounted present value of-the opportunities that would be
foregone. If all of the land to be used in the development is to be
acquired on the open market, then the market price of the land, which
will be included as part of the development's construction cost provides
a reasonable estimate of the value of these foregone opportunities. In
this instance the opportunity costs have already been-counted once (as
construction costs) so it would be incorrect@ to count them again.

If the city presently owns some of the land that will be used in the
development, however, and it is contributed free-of-charge or at a
below-market price to the development, then true construction costs will
be understated. Consequently, if city-owned land is to be used in the
lakefront development, the analyst must determine the market value of
that land. This market value should be used as a measure of the social
cost of using city-owned land in the development of the lakefront.

10.5 DISPLAY OF SUMHARY BENEFIT-COST MEASURES

The next step in the writing of a benefit-cost report is the preparation
of a display summarizing the benefit-cost measures, either benefit-cost
ratios or"net present values. The purpose of this display is to convey
to the reader (the policymaker) the concise information regarding the
project's benefits and costs, and perhaps information about the distri-
bution of those benefits and costs. The display should also contain
the results of the more important sensitivity analyses, to aid policy-
makers in understanding the study's more crucial assumptions.

Table 10-1 is an example of such a display. This table lists the summary
benefit-cost measures resulting from sensitivity analyses involving:
(a) the estimated number of beneficiaries (denoted here simply by x, y,
and z), (b) the benefits per beneficiary (represented by B and B'),
(c) the discount rate (4 percent and 10 percent are used in this illus-
tration), and (d) the time period (50 and 100 years are merely illustra-
tive).

Completing such a table quickly reveals the more important assumptions.
For instance, it may indicate that the benefit-cost estimates are highly
sensitive to the choice of the discount rate, but relatively insensitive
to changes in the time period. This would suggest that policymakers
spend more time-determining which discount rate best suit's the needs
of the City. The table may also highlight the need for more precise
estimates.

One problem posed by a table such as Table 10-1 is how to interpret a
conflicting set of benefit-cost statistics, some suggesting that the
project is efficient and some suggesting it is inefficient. One approach
to this problem is to. ignore the conflicting implications and select that
set of assumptions that appear most reasonable or appropriate. For in-
stance, if a policymaker feels that public projects of this sort should
be evaluated with low discount rates, the fact that, say, relatively
high discount rates tend to make the project appear inefficient is of
little importance. While the analyst is free to select the assumptions
that appear the most appropriate, he should also present the results of
the sensitivity analyses to provide policymakers the data for making
their own judgments and recommendations.

10.6 INTANGIBLE EFFECTS

Presentation of the measured benefits and costs should not be allowed to
obscure the fact that there may be substantial unmeasured or intangible
effects. The presentation should be carefully qualified with the in-
formation that some benefits or costs are unmeasured.

One method for bringing these omitted factors into the evaluation,
aside from merely mentioning them, is to ask the following question:
Could the unmeasured benefits (costs) plausibly be large enough to make
an unfavorable benefit-cost ratio favorable (or vice versa)? In the
lakefront development project, the relevant question would be: Could
the unmeasured net benefits of the project (per beneficihry) change the
favorability of net cost (where net cost is the difference between the
measured social benefits of the project and the measured social costs)?

Table 10-2 shows how the estimates could be d4.sp:laved to allow for alter-
native assumptions regarding, say, the anticipated number of beneficiaries,
discount rates, time periods, and average measured benefits per beneficiary.
This table would have to be completed only for those assumptions that
yielded benefit-cost ratios less than unity (assuming that all costs had

-70

Table 10-1

Alternative Benefit-Cost Rating for Chicago's
Lakefront Development Plan

Number of beneficiaries and benefit measures used
lox" Beneficiaries "y" Beneficiaries 'Y' Beneficiaries
Average benefits Average benefits Average benefits
Discount rate per beneficiary per beneficiary per beneficiary
and time period B B1 B B1 B B1

4 percent rate

50 years

100 years

10 percent rate

50 years

100 years

Table 10-2

Estimate of the New Measured Cost Per Beneficiary
Provided by Lakefront Development

Number of beneficiaries and benefit measures used.
FIX" Beneficiaries "y" Benefi iaries "z" Beneficiaries
Average benefits Average benefits Average benefits
Discount rate per beneficiary per beneficiary per beneficiary
and time period B B1 B B1 B B1

4 percent rate

50 years

100 years

10 percent rate

50 years

100 years

been measured with an acceptable degree of accuracy). If this analysis
reveals, for instance, that intangible benefits would only have to be
$5 to $25 per beneficiary in order to alter all the unfavorable beriefit-
cost ratios, policymakers could feel confident in approving the@@-,,Arcject
on efficiency grounds; on the other hand, if the intangible benefits
would have to be as large as $1,000 to $10,000 per beneficiary before
the project would appear efficient, policymakers would be much more
skeptical about the project's desirability. With information of this
sort, policymakers should find it easier to come to grips with the prob-
lems created by intangibles.

10.7 DISTRIBUTION OF BENEFITS AND COSTS 0@w

In addition to the efficiency considerations, policymakers may find it
useful to have information about how the project's benefits and costs
are distributed across different groups of people. In some instances,
the distributional considerations may be of overriding importance from
a policy standpoint and the analyst is reminded that this information
may be needed in the display. '(See Section 9.)

BIBLIOGRAPHY

SECTION 11

SELECTED BIBLIOGRAPHY

Arrow, K.J., and Levhari, D. 1969. Uniqueness of the Internal Rate of
Return with Variable Life of Investment. Economic Journal,
79: 560-566.

Arrow, K.J., and Lind, R.C. 1970. Uncertainty and the Evaluation of
Public Investment Decisions. American Economic Review,
60: 364-378.

Bailey, M.J. 1959.,Formal Criteria for Investment Decisions.
Journal of Political Economy, 67: 476-488.

Barrett, L.B. and T..E. Waddell. Cost of Air Pollution Damage: A
Status Report. U.S. EPA Pub. No. AP-85. Research Triangle Park,
N.C., 1973.

Baumol, W.J. 1968. On the Social Rate of Discount-American Economic
Review, 58: 788-802.

Broadway, Robin. 1976. Integrating Equity and Efficiency in Applied
Welfare Economics. Quarterly Journal of Economics, 90: 541-556.

Burt, O.R. and D. Brewer, "Estimation of Net Social Benefits from
Outdoor Recreation," Econometrica, 39.5(813-827) September, 1971.

Cesario, F.J. 1976. Value of Time in Recreation Benefit Studies.
Land Economics, 52: 32-41.

Clawson, M., and Knetsch, J.L. 1971. Economics of Outdoor Recreation.
Baltimore: Johns Hopkins Press.

Common, M.S. 1973. A Note on the Use of the Clawson Method for the
Evaluation of Recreation Site Benefits. Regional Studies,
7: 401-406.

Coomber, N.H., and Biswas, A.K. 1972. Evaluation of Environmental
Intangibles. Bronxville, N.Y.: Genera Press.

Currie, J.M.; Murphy, J.A.; and Schmitz, A. 1971. The Concept of
Economic Surplus and its Use in Economic Analysis. Economic
Journal, 81: 741-799.

Dasgupta, P.; Sen, A.; and Marglin, S. 1972. Guidelines for Project
Evaluation. New York: United Nations.

Dasgupta, A.K., and Pearce, D.W. 1972. Cost-Benefit Analysis: Theory
and Practice. London: Macmillan & Co.

Darling, Arthur H. 1973. Measuring Beri61-1' s Generated by Urban Water
Parks. Land Economics, 49: 22-34.

Davidson, P.; Adams, F.G.; and Seneca, J. 1966. The Social Value of
Water Recreational Facilities Resulting from an Improvement in
Water Qaulity: The Delaware Estuary. In Water Research, ed. r
A Kneese and S.C. Smith. Baltimore: Johns Hopkins Press.

Dorfman, R., ed. 1965. Measuring the B nefit of Governmental
Investments. Washington, D.C.: Brookings Institution.

Eckstein, 0. 1958. Water Resources Development. Cambridge, Mass.:
Harvard University Press.

Federal Water Pollution Control Administration. Delaware Estuary
Comprehensive Study: Preliminary Report and Findings.
Philadelphia, July, 1966.

Feldstein, M. 1964a. Net Social Benefit Calculation and the Public
Investment Decision. Oxford Economic Papera, 16: 114-131.

Feldstein, M. 1964b. The Social Time Preference Discount Rate in Cost
Benefit Analysis. Economic Journal, 74: 360-379.

Feldstein, M.S., and Flemming, J.S. 1964. The Problem of Time Stream
Evaluation: Present Value Versus Internal Rate of Return Rules.
Bulletin of Oxford University Institute of Economics and
Statistics, 26: 79-85.

Fischer, David W., Willingness to Pay as a Behavioral Criterion for-
Environmental Decision-Making, Journal of Environmental
Management, 3(29-41) 1975.

Harberger, Arnold C., et. al., eds. 1972. Benefit Cost Analysis 1971:
An Aldine Annual. Chicago: Aldine.

Harrison, A.J., and Ouarmby, D.A. 1974. The Value of Time. In
Cost Benefit Analysis, ed. R. Layard. Baltimore: Penguin.

Haveman, Robert H., et. al.,, eds. 1974. Benefit Cost and Policy
Analysis 1973; An Aldine Annual. Chicago: Aldine.

Haveman, R.H., and Krutilla, J.V. 1958. Unemployment, Idle Capacity
and the Evaluation of Public Expenditure. Washington, D.C.:
Resources for the Future, Inc.

Havighurst, C.C. A Survey of Air Pollution Litigation in the
Philadelphia Area. Durham, North Carolina: U.S. DHEW, PHS,
National Air Pollution Control Administration (Final Report,
Contract No. CPA-22-69-11ZY,.'4-,December, 1969.

Howe, C.W.,1,971..Benefit-Cost Analysis for Water System Planning.
Water Resources Monograph, 2. Washington, D.C.: American
Geophysical Union.

James, E. 1969. On the Social Rate of Discount: Comment. American
Economic Review, 59: 912-916.

Knetsch, Jack L., and Robert.K. Davis. 1966. Comparison of Methods for
Recreation Evaluation. Water Research, ed. by Allen V. Kneese and
Stephen C. Smith. Baltimore, The Johns Hopkins University Press.

Krutilla, J., and Cicchetti, C. 1972. Evaluating Benefits of
Environmental Resources with Special Application to the Hells
Canyon. Natural Resources Journal, 12: 1-29.

Krutilla, J.V., and Eckstein, 0. 1958. Multiple Purpose River
Development. Studies in Applied Economic Analysis. Baltimore:
Johns Hopkins University Press.

Lesourne, Jacques. 1975. Cost-Benefit Analysis and Economic Theory.
New York: American Elsevier.

Lind, R.C. 1964. The Social Rate of Discount and the Optimal Rate of
Investment: Further Comment. Quarterly Journal of Economics,
78: 336-345.

Maass, A., et. al. 1962. Design of Water Resources Systems.
New York: Macmillan.

Maass, A. 1966. Benefit-Cost Analysis: Its Relevance to Public
Investment Decisions. Quarterly Journal of Economics, 80: 208-226.

Maler, K.G., and Wyzga, R.E. 1976. Economic Measurement of Environmental
Damage. Paris: Organization for Economic Cooperation and
Development.

Mansfield, N.W. 1971. The Estimation of Benefits from Recreation Sites
and the Provision of a New Recreation Facility. Regional Studies,
5: 55-69.

Marglin, S.A. 1963a. The Social Rate of Discount and Optimal Rate of
Investment. Quarterly Journal of Economics, 77: 95-111.

Marglin, S.A. 1963b. The Opportunity Costs of Public Investment.
Quarterly Journal of Economics, 77: 274-289.

Marglin, S.A. 1967. Public Investment Criteria. Cambridge, Mass.: MIT
Press.

McKean, N. 1968. The Use of Shadow Prices. In Problems in Public
Expenditure Analysi@i, ed. S.B. Chase, pp. 33-77. Washington, D.C.:
Brookings Institution.

Merewitz, L. 1966. Recreation Benefits of Water Resource Development.
Water Resources Research, 2: 625-640.

Musgrave, R. and Musgrave, P. 1976. Public Finance in Theory and
Practice. New York: McGraw-Hill.

National Academy of Sciences. 1975. Asse'ssing.Demand for Outdoor
Recreation. U.S. Government Printing Office Stock Number
624-016-00081-1. Washington, D.C.

Niskanen, William A., et. al., eds. 1973. Benefit Cost and Policy
Analysis 1972: An Aldine Annual. Chicago: Aldine.

Pauly, M.V. 1970. Risk and the Social Rate of Discount. American
Economic Review, 60: 195-198.

Pearse, Peter H. 1968. A New Appro;7i,:h to the Evaluation of Nonpriced
Recreational Resources. Land ---Iconomics, 44: 87-99.

Plager, S.J., R.W. Findley, M. Frankel, P. Loquercio, J.S. Moore, and
P.D. Schomer., Draft. Economic Study of Proposed Motor Vehicle
Noise Regulations. Illinois Task Force on Noise, IIEQ (xerox copy)
April 16,'1976.

Prest, A.R., and Turvey, R. 1965. Cost-Benefit Analysis: A Survey.
Economic Journal, 75: 683-735.

Reiling, S.D., K.C. Gibbs and H.H. Stoerener. Economic Benefits from
an Improvement in Water Quali@yb Washington: EPA. January, 1973.

Schramm, G. 1973. Accounting for Non-economic Gaols in Benefit-Cost
Analysis. Journal of Environmental Management, 1: 129-150.

Stanford Research Institute. Economic Impact of Air Pollutants on Plants..
(CRC-APRAC Project No. CAPCA-2-68).Palo Alto, California, Annual
Report. August, 1970.

Trice, A.H., and Wood, S.E. 1958. Measurement of Recreation Benefits.
Land Economics, 34: 196-207.

Turvey, R. 1963. On Divergence between Social Cost and Private Cost.
Economica, 30: 309-313.

U.S. Congress, Subcommittee on Economy in Government of the Joint
Economic Committee. 1967. The Planning-Programmin@-Bud geting
System:.Progress and Potentials. Hearings. 90th Cong., 2nd sess,
'July 30 and 31 and August 1, 1968 (Washington, D.C.: Government
Printing Office).

U.S. Congress. 1968a. Economic AnalXsis of Public Investment Decisions:
Interest Rate Policy and Discounting Analysis. Hearings. 90th Cong.,
1st sess., September 14, 19, 20, and 21, 1976 (Washington, D.C.:
Government Printing Office).

U.S. Congress. 1968b. Interest Rate Guidelines for Federal Decisionmaking.
Hearings. 90th Cong., 2nd sess., January 29, 1,968 (Washington, D.C.:
Government Printing Office).

Usher,, Dan. 1964. The Social Rate of Discount and the Optimal Rate
of Investment: Comment. Quarterly Journal of Economics, 78: 641-644.

Weisbrod, B. 1968. Income Redistribution Effects and Benefit-Cost
Analysis. In Problems in Public Expenditure Analysis, ed.
B.S. Chase. Washington, D.C.: Brookings Insitution.

3 6668 14104 0925

r
ITER
o N